This is an extraordinary, beautiful--and excessive--book that places certain a priori requirements on the reader. It is not a book for the normal person.
You have to be a lover of languages, as well as a student of all the weird things your brain and ego do to you when you learn a new language. It will help you to enjoy this book if you speak some French as well as a little Italian. You'll have to be a lover of poetry, including poetry's various technical elements. You have to be able to appreciate wordplay, a good turn of phrase, and other geeky nuances of language. You have to enjoy fooling around with words!
If that's not you, this is not your book. And even if it is you, you'll still need to be extremely patient with the author. More on that in a moment.
Just like with this author's beautiful and better-known Godel, Escher, Bach, it is next to impossible to describe this book. Literally, Le Ton beau de Marot is about a subtle little poem, "Ma Mignonne" by the minor French poet Clement Marot, and the various challenges author Douglas Hofstadter faces when he attempts to translate it into English. Do you keep all the poem's structural elements in your translation? Should your translation rhyme in the new language? And if so, what kinds of semantic liberties must you take in order to adhere to these structural elements?
[A quick affiliate link to Amazon for those readers who would like to support my work here: if you purchase your Amazon products via any affiliate link from this site, or from my sister site Casual Kitchen, I will receive a small affiliate commission at no extra cost to you. Thank you!]
It sounds abstruse already. But over the course of exploring these various translation issues and teasing apart the underlying ideas at play, the author delves into an unimaginably wide range of topics. A patient reader emerges from this book with a tremendous reading and media list of authors, poets, music, plays, even films, to explore. I'm already grateful to this book for sending me to Alexander Pushkin's Eugene Onegin, a work discussed in great depth in Chapters 8 and 9, two of this book's finest chapters. Thus this book has given me a profound gift: years' worth of literary works to read and ransack for ideas.
And, somehow, the author also makes this long and winding book into a beautiful love letter to his wife, who tragically died young.
In between each of the book's chapters are short interludes where the author shares all kinds of translations of Marot's poem: some maintaining extreme fidelity to the poem's form and structure, others taking striking, even shocking, liberties.
That's the literal level of what this book is about. Perhaps a metaphor to consider when trying to describe Le Ton beau de Marot is to think about how different people experience a good movie. You can of course simply enjoy the movie. But you can also enjoy all of what goes into making the movie: the various layers of artifice employed, the shot choices, the special effects, the music. And you can be a student of whether these things are used blatantly, subtly--or ideally, invisibly--to create the viewing experience.
Those who enjoy thinking about these layers "behind" a work--in firm, music, literature, poetry or any artistic domain--will enjoy this book too.
All that said, there are a few sections in Le Ton beau that will fatigue even the most patient reader. See for example Chapter 7, which contains a few too many Tom Lehrer songs followed by a multipage lecture on the deplorable sexism of the phrase "you guys." Likewise, Chapters 11 and 12 should be tightened as the author becomes a bit masturbatory in his various musings about the nature of translation. And Chapters 14 and 15 really lose their way as the author explores the untranslatable and un-understandable.
This book is an unbelievably generous gift to a certain type of reader. Yes, you'll have to grind through a few chapters that could have been cut. But it's well worth it.
For language geeks, pair with:
William Empson: Seven Types Of Ambiguity
George Lakoff and Mark Johnson: Metaphors We Live By
[Links take you to reviews on this site]
[Readers, a severe warning: the notes that follow below are interminable. If you are at all doubtful as to whether this book is for you, read no further. Spare yourselves! If, however, you really think this book might interest you, start by skimming the bolded parts below. Also, feel free to look over the gigantic reading list and media list at the very bottom of the post.]
Notes:
Introduction: In Joy and in Sorrow
xiiiff "This long book was sparked almost ten years ago by my trying to translate one sweet, old, small, elegant French poem into English, and by the unexpected snowballing of that one first try into a mammoth avalanche of further translations by old friends, new friends, relatives, colleagues, students, strangers, and, not least, myself." Comments here on the exuberance a writer gets from creating a deeply-integrated work; the author says the last time he feels like he did anything really similar was twenty years earlier when writing Godel, Escher, Bach. "If one is lucky, one has the luxury of becoming totally immersed in an artistic project, letting almost all other things go by the wayside--family, friends, students, colleagues, food, bills, correspondence, neatness, books, music, movies, shopping, and sleep, to give a few examples." Interesting musings here on both his process of writing the full first draft, then "wrapping around" and cycling through the whole thing all over again, and how over the course of this second cycle through a "global tone" emerges: things come to him, ideas crystallize that unify the whole work, what he calls a sort of convergence to a global tone. Also on how he can't believe how much rewriting he does, he considers dozens of variants of each sentence.
xviff Contrasting his writing style from Stephen J. Gould, who writes from an outline and a strict plan, Hofstadter instead depends on the appearance of unexpected ideas and coincidences. "To me, that is the entire excitement and beauty of the creative process. Things come from out of left field and are drawn in and integrated and then become central."
xviii "I think one strongly needs deadlines of some sort or other in life; they act as organizers without which one would simply flail about for unlimited amounts of time." [Holy cow is this ever true.]
xviiiff Discussion here of how the publisher gave him total control over the book, from the cover art to the typefaces, to even the kerning--the spaces between characters on the page. The author muses that instead of him enjoying total control over the book, the book had total control over him as he was forced to rewrite and rewrite passages so that a given page boundary would come out exactly where he wanted, and also to make sure no poem was ever broken across page boundaries; also comments here on how "constraints" like this can be a source of creativity--just like the constraints of poetic meter or of translation can drive creativity. Thus the book turns out to have interlocking constraints: the author walks through an example of a choice of a specific letter at the beginning of Chapter 2 led to a whole series of changes to that chapter, including changes to the very ideas expressed.
xxiv Finally, the author dedicates the book to his wife who died suddenly in 1993 of a glioblastoma, she was 42 and they had been married just 8 years, with two young children.
Chapter 1: The Life and Rhymes of Clément Marot
1ff This chapter is brief history of the poet himself, Clément Marot, and it's written in kind of a prosody form, there's a rhythm and scanning to the text and occasional poetic alliteration and rhyming; on the poet's style, mostly light verse, sparked with joy filled with wordplay and "phonetic regroupings" like lave oye, la voie, l'avoie.
3 "As fifteen thirty-four was just opening its door, with deep drifts on the ground giving grounds to stay in..." [This gives a good sense of the vibe of this first chapter. I'm not sure it works, but it is cute and witty, and it probably was incredibly difficult and time-consuming to structure sentences like this throughout the chapter.] Note that Marot, like many in his day, got caught up in the sectarian conflict throughout France between Protestants and Catholics; he had his property taken, and he had to escape: first to Navarre, then to Italy and Switzerland. Two years later, he was able to return to France, but right then, in 1542, an anti-Lutheran movement exploded in France, and he had to flee the country again, first to Geneva, then eventually to Piedmont, and then to Turin where he died in 1544 at age 47. Various puns in French throughout this entire section, I'll share one here, regarding "le tombeau de Marot" (Marot's gravestone/tombsite), "le tome beau de Marot" (the great book of Marot), and "le ton beau de Marot" (Marot's graceful/beautiful tone or good style).
Poems I: Original and Literal
1a On "Ma Mignonne," a poem Clément Marot wrote to Jeanne d'Albret de Navarre when she was sick as a sort of get-well card:
[Note that this is actually readable with a middling level of competence in French and an occasional peek at Google translate.]
1aff The author describes various structural and syntactical features; he talks about when he sent this poem off to various friends along with a list of the poem's formal properties that he felt were crucial to consider in any translation. "I always supplied this list, obvious though its terms were, because I did not want them to be overlooked. If someone knowingly chose to disrespect one or another of them, that would be all right--or at least far more justifiable, in my opinion, then doing so out of ignorance." [Here we get some of our first inklings of one of the author's foundational ideas: on the critical importance of maintaining certain elements of form in any act of translation. In fact, literally, the medium is the message here in the sense that the form carries quite a lot of the information of the message itself.] Ironically the author himself is shocked to learn that there are still more and formal properties of the poem that he himself had not noticed, "one of them extremely beautiful and yet quite subtle to spot, although after the fact it seems obvious." [The answer here--I think--is the semantic couplets and the rhyming couplets are out of phase throughout the poem: the semantic couplets do not rhyme (Qui se couche/En danger) and the rhyming couplets are not semantically linked, rather they are part of semantically separate phrases (En danger/Pour monger). See below, in Poems III: Antique Airs, for more on this.]
2aff What follows here is a series of English translations of this poem, starting with a bland literal version, which functions as a "crib" to explain line by line what the poem "means." Note the transition from "you" to "thou" here which reflects the transition from vous to tu in French, but "thou" fails to capture the same emotional content of tu. Also on the problem of the first line: in French you know it's addressed to a woman ("mignonne") but in English we don't have gendered words like this. And so how can you convey the idea that the my cute or my sweet is a woman? [Or is it implied by the title "To a Sick Damsel"?] And then how would you hold all these aspects in place while keeping to the two syllables per line structure, maintaining the proper syllabic stresses, etc? "Exactly what kinds of things are we supposed to pay attention to, and what kinds of things are we free to ignore?"
3a Next is a version that spends more time imitating and replicating the poem's form and scansion, thus the poem flows far more smoothly, but at the expense of full literality.
4a Next is a translation [which kind of reminds you of a computer program with comments] where the author gives multiple meanings for each word along with context like (feminine); the author calls it "a legitimate exploration--albeit a rather extreme one--in the art of literal translation." He says it reminds him of modern buildings that have their plumbing and wiring visible instead of hidden behind walls, while it gives far more context to non-French speakers of the various nuances of the poem:
My sweet/cute [one] (feminine),
I [to] you (respectful) give/bid/convey
The good day (i.e., a hello, i.e., greetings).
4a The author has the reader imagine that you were trying to find the "best" English translation: what criteria would you use, which translation would you pick? Uou might find some translations daring or graceful, but why? And why were they yet not the best translation? Also, would you feel free to tamper with a near-perfect translation done by someone else? "Thinking about such issues will surely carry you deep into the spirit of this book."
5a Finally, two other translations: one which is sort of a machine-translated paragraph without line breaks or stresses, and finally a "touchstone version" which the author created to jog his memory and think of the chain of "keys" to the message. [Again it's worth noting that he's writing these commentary passages next to each poem version with rhyme, stress and poetic prosody!]
The author then asks the reader to do your own attempt at translation. Here's my own version: an alert reader should see a simple syllabic pattern.]
My love,
I wish you
Good day;
Your stay-in:
Prison.
To good health
Return.
You open
Your door
And can leave
Right now:
Your Daniel
Says so!
Gourmande,
So your health
Won't be
In danger,
Go eat
Confitures.
If you
Stay indoors,
You'll fade,
And you'll pale,
And lose
Too much weight!
God grant
You good health,
My love.
Chapter 2: For the Love of a Poem from Days Long, Long Gone
5ff The author describes how he discovered the poem "Ma Mignonne," fell in love with it, and memorized it while he was in college: "So taken was I with 'Ma Mignonne', so intoxicated by its miniature lines and its ultradensely packed-in rhymes, that I was unable to resist committing it to memory, where it stayed fully intact for twenty-five years and then some, and from which it was stochastically resuscitated, perhaps two or three times a decade for just a day or two, after which it would plunge back into deep dormancy for another few winters." [He's absolutely right about the events that happen after you memorize a poem: I've done this with certain English language poems and they really do stay there, you do pull them up from time to time and then over time your memory of it becomes patchier and patchier. I recommend memorizing poetry, in particular a poem that really resonates with you. It's a good exercise on a number of levels, and it really teaches you the subtleties and nuances of good poetry.] He then recites it for his wife while he's working on translations of his book Godel, Escher, Bach; suddenly he sees the poem in a new light as "a thing to translate," shocked that he had never thought of the work in this way before.
6ff He finally works out a translated version that he was proud of; but then it leads to a cascade of intellectual activity as this book grew around it. He sends the poem, along with his description of the structural elements of it as well as his two "literal translation" (we saw them in Poems I) to 50 or 100 friends, challenging his friends to make "an artistic equivalent" of Marot's poem in their own language in the face of the various "mutual vying pressures" of the poem's various constraints and structural elements.
8 He writes in this letter things like translators will be "pushed to the very fringes of their native language"; questions on what constraints you might loosen, what liberties you might take, does "taking liberties" get to a point where you become sloppy? And then he writes: "I must say, I have really had a ball translating 'Ma Mignonne' and thinking about these issues. Once I had done one translation, thus proving to myself that it was feasible, somehow I was able to do others with much greater ease than I did my first one. I am fascinated by this effect, and don't fully understand it."
8 "...deep issues emerge most clearly, and are therefore best studied, in well-chosen microcosms." [This is an interesting quote that sounds like it could apply very widely to a lot of domains...]
9 [Good restating of the overall theme of this book here in this quote]: "In rereading this old letter after some years had passed, I was impressed by the way it anticipated so many issues that recur and dance together throughout this book, like themes in a sprawling counterpoint: content versus form, constraints, trade-offs, slippability, strictness and laxness, the creative process, drafts, revisions, breakthroughs, abstract style, varieties of literality, tradition pushed to its limits, effects of the target language, the mystery of how one success builds on another..."
9 Also comments here on the "rickety bridge" effect, where the first spanning of a translation chasm leads to stronger and subtler bridges.
9ff His letter produced all kinds of responses from friends, friends of friends who became friends, etc. [It's a neat idea to send out something like this.] He uses it as a basis for lecture talks that he was giving on translation, and this led to more input from people on even more nuances of translation issues, as well as new versions of translations of the poem. Then he has the idea that this "might make a cute little book." It started as book designed to accompany a lecture series he was hired to do, but then grew into a much larger project.
12 Cute comment here about how huge the ultimate project turned out to be; had he known he would not have the courage to do it! "Luckily, though, I do not have a gift of clairvoyance, and so I mostly plunged ahead..." [Ignorance really can be courage sometimes...] The author also ties in his wife more and more throughout the book because she passed away around the same time.
13ff Comments here on another work: a collection of translations of Horace's Ode to Pyrrha that is quite similar in that many people took wild liberties and imposed all kinds of interesting structures; the author asks "And yet are we perceiving Horace, or is his image too distorted by these refracting prisms? Such questions will hound us as we now turn back to Clément Marot and his equally charming and challenging minuscule opus..."
Poems II: Gals and Trysts
6a The author shares his very first translation of "Ma Mignonne." [It reads a little bit like a gift card to be honest but it's better than anything I could do. It adheres loosely to the literality of the poem, while still maintaining the meter and rhyming couplet aspects.] The author calls this act of translation "an intense little burst of joy"; he notes that he had tried weeks earlier and given up in despair, but then came back to it in a fresh frame of mind and was willing to play with it more. He refers to talking about butter rather than jam, and lays out a pretty impressive pun here: "Did following this buttery trail constitute utter betrayal?"
6-7 Also an interesting comment on the word "jail" and "jailbreak": in the first version here on page 6b jail is structured as a one-syllable word, but in the version on 7b the word jailbreak is a three-syllable word, using "jail" as a diphthong in order to meet the metrical requirements.
8aff The author receives two separate versions from his friend Bob French. who was a professional French to English translator. It was nothing what he expected: way different from anything he had even considered, and it made him depressed about his own lighthearted efforts; he even becomes embarrassed by his exclamation points. [This is so beautifully honest: you see what someone else does when they do it well and it makes you ashamed of what you've done by comparison. It's human nature.] Comments on how Bob French was faithful to the elegance of the epoch in which the poem was written, yet another translation aspect that never crossed Hofstadter's mind. Finally here debate about line count: he didn't realize it at first but French's version had 30 lines, not 28, and they have a aesthetic debate about whether the line count really even matters (per French it doesn't as long as you're in the same ballpark), but the number of syllables per line and other metrical aspects really do matter.
8b The first version that Robert French wrote is quite beautiful and evocative, although it has sections that don't work well: see lines 24-26 "You'll put on / But then cede / Pounds you need."
9a Hofstadter discovers a translation error in French's poem, as well as a possible fix to correct both the semantic problem and the line count. The author also comments about a sort of "meaning layer" in the poem that exists in the original: that it could be written to a lover or a mistress, there was more ambiguity in the original poem than Hofstadter initially realized.
9a "How far a translator can reasonably drift from a literal text has everything to do with the fabric of human associations--with what lies mentally close to what, and what lies far away. Such associations come, of course, from deep familiarity with how the world itself is structured. If one has lived through millions of complex experiences, as we all have--including vicarious ones, from books we've read to movies we've seen to adventures we've heard friends relate--then just a few words can trigger rich imagery at a conscious level, as well as vast clouds of associations at a more subliminal, invisible level."
Chapter 3: How Jolly the Lot of the Oligoglot
15ff [This chapter will resonate with readers who have had experience with language learning; I think it might resonate with people curious about language learning; but for those readers uninterested in all the wacky things that happen to your mind when you learn and think in other languages, I think this chapter will give you an early cue for whether it's worth continuing to read this book.] The author, now a teenager, begins learning French in preparation for his father's sabbatical in Geneva; he has a knack for the accent but he hides it in class in school in America. How he becomes a fanatic about good pronunciation and also becomes fascinated by "what it means to think in another language"; also comments here on how, later, after he had studied many languages to varying levels of competence, he talks about how his level fluctuates due to "hugely and unpredictably" random things: mood, topic, his conversation partner, and other reasons he can't even understand. On the naive and meaningless question of "how many languages do you speak?" which depends what you mean by "speak." Note the authors neologism "pilingual" which means "I speak "pi" languages because the "fractional fluency" I have in all the languages I've learned adds up to pi. [Cute.]
17 [An offhand comment on the author's experience learning "what at the time was called Serbo-Croation"... which sent this reader down a rabbit hole to learn about this phrase being invented by the Yugoslav dictatorship in the 20th century to promote national unity; in reality there is a family of closely related languages but with distinct national and cultural differences that are today far more emphasized; it was an invented catch-all term for a cluster of languages now known as Bosnian, Croatian, Montenegrin, and Serbian. If you think about this a little bit, you realize that the underlying "language sovereignty" is everything: this is why in Catalunya they make sure that everyone speaks Catalan first and Spanish second, language is a huge component of what makes a "people"; also it's worth noting that the concept of a "people" is quite different from a "nation" in the modern nation-state sense, which is a piece of paper and land.]
17 Comments here also on how one is never satisfied with one's language fluency in any language, ever, although you do get occasional moments of joy and confidence when you surprise yourself. [Holy cow is this ever true.]
17ff He has some striking intellectual moments in his French grammar class: see for example when he learns the conditional tense and comes to understand the idea of "would-ness," which let him "see into" the word "would" in English. [Basically he learned about his own language by learning another one, which is something that happened to me as well]. And he started to understand certain elements of the logic behind grammar, seeing it as sort of a formal logic behind languages; also an appreciation for "word boundaries" [which are particularly blurry in French! But also in English too depending on one's enunciation], and other subtleties of understanding native spoken expressions, which are often slangy, slurred, smushed together, etc. [I am deeply grateful to a close Quebecoise friend who taught me all the good church words, as well as taught me to stop carefully sounding out phrases like "Il n'y a pas de quoi" and "il y a" and instead saying "padquoi" and "ya"].
18ff He learns there there are levels to things [everybody to their horror--or to their excitement--learns that there are levels to things, and that they're nowhere near the level they thought they were at, and nowhere near a level of true expertise. Language learning is humbling, in a good way. In the following paragraphs here the author really captures some of the crushing lows and wonderful highs that come with language learning as you move from level to level, as you realize how little you really know and how much more there is to know; as you learn how much you'll never know. Some people are really egoically broken by this and don't even make a real attempt, others survive the ego death and come out the other side stronger and more psychologically robust.]
21 The author keeps a notebook where he faithfully jots down all of his cross-language errors in order to understand how language works in a human mind, and what "thinking in French" really means.
22 The author tosses off a good list of beautiful French singers here for anyone looking for good French music:
Charles Trenet
Jacqueline François
Patachou
Line Renaud
Isabelle Aubret
Lucienne Delyle
Renée Lebas
Yves Montand
Brigitte Bardot [I had no idea she sang, the author says "and she was good!"]
Léo Ferré
Georges Brassens
22ff He decides to learn German and goes through some of the funny compound words that German has, many of which are illogical, but many of which let you figure out exactly what the word means by working out the compound word. And then this word: Agrarstrukturverbesserungsmaßn
or "Measures taken to improve the agricultural structure"
26ff On certain cognates in German like brustwarze which is a compound word meaning "nipple" but literally meaning "breast-wart"; also the word reisschleim which is a compound word meaning "rice-slime"; somehow Germans use these words and don't have the same corresponding revulsion towards the word "wart" or "slime" that we have as English speakers; somehow these compound words in German rob them of the negativity that they would have otherwise.
28ff He becomes entranced by Italian and decides he has to learn it, and he loved the idea that he was able to parlay his French into Italian. [He's absolutely right, if you have a decent level of French you can pretty much read Italian, lots of verbs translate pretty directly, etc.]
30 Another very interesting comment here on when he studied Italian in college, he was assigned literary criticism essays--which he hated-writing--but he would make them fun by leafing through Italian magazines and copying phrases and snippets that sounded quintessentially Italian, and then use them as a sort of structure to build his essay around. This is sort of a method of being guided by form better than content and it turned the writing task somewhat upside down, but since then he's actually adopted this technique with writing in English, setting the structure first and then letting the words and ideas in his head fit the mold that he's created or fit the constraints that he's created. [Great idea!]
32 Comments on some of the music he learned during a lecture tour in Italy with his wife who he had just married: he mentions the songs of the 1930s and 1940s of Italy as a sort of Golden Era of Song comparable to our Bing Crosby and Jerome Kern era, but nobody remembers these artists today [either in Italy or in the United States]: "To me, there is something extraordinarily sad, even frightening, about this contemporary trend, on both sides of the Atlantic, towards near-total detachment from an unbelievably glorious musical past." [This is a deeply sobering quote, and it shows how the system wants you to forget your past: it's "year zero" now, we have to replace the past with whatever dreck the media wants to force on us now.]
35ff He then builds Spanish on his tower of languages, on top of his Italian, French and English. He speaks about how you have a matrix of words across these languages, e.g.: saber/sapere/savoir and also how there's various bizarre cross-linguistic contaminations that happen across the various languages that sit in your brain. [Again, he's absolutely right: you can learn new verbs in a new language much more rapidly--at least across the Romance family of languages--by connecting French, Spanish, Italian and English. Of course once in a while this creates astoundingly hilarious mistakes, hilarious for everyone around you except for you! Like the time I wanted to convey my enthusiasm to a group of Chileans and said, regrettably, "Estoy exitado!"]
40ff On accents and the idea of "abstracting away the accent layer": can an accent exist in say Chinese characters drawn by an American, a type of "visual accent"; also on the idea of "accent" in the way baseball or chess is played by people of different cultures; or even abstracted away to the computer programming language that the author uses: see for example the fact that "Algol" was his first computer language, would that mean he has an "Algol accent" when he programs in Lisp? Do physicists, many of whom move over to molecular biology, have a "physics accent"? Do Americans drive with an "American accent" when they're driving in Europe? Or does a male writer have a "masculine accent" and vice versa?
44 Cute brief blurb here about O.B. Hardison who invited Hofstadter to a lecture series on translation long before he knew anything about the domain; he was baffled at why he was invited, but Hardison told him, "As soon as I read your book Godel, Escher, Bach I was sure you would be the perfect person to invite" because his book is all about translation; the author had no idea about considering his own book in this light. "...he had pointed out something to me about my book that I myself had not seen, and for this I have always felt truly appreciative."
47 Discussion here of various examples of analogical transfer: from language to language: one language gives you low-level competence in another one; from one type of puzzle you can borrow techniques and solve other types of puzzles; from one sport you get confidence in another: see for example ice skating helps you with rollerblading and even skiing, etc. The author even uses examples in social interactions where you use certain types of banter with new people and accelerate getting to know people.
49ff On moving up in communication "freedom" based on your facility with various languages. THe author describes an experience in Poland, speaking with someone in his very limited Polish, but finds that the colleague speaks German, so click, he can raise his communication level. Then he talks to someone else later in French, so click, up another level; then later he speaks in English, where he has no limitations. He likens it to being released into a bigger and then bigger cell, with fewer and fewer cognitive shackles, as he moved into each more expansive language. "...in English I no longer felt the walls of my cell..."
50ff Long discussion here on the works of scifi writer Stanislaw Lem and the author's exposure to his writing all of Lem's wordplay and neologisms. [This section introduced me to some very good novels by Stanislaw Lem, including Solaris, Fiasco and The Cyberiad.] On various "cowardly capitulations" a translator could use to try to stay "true" to the original.
56 Note in particular here where the author goes through one particular "cowardly capitulation" which is to use a literal translation of a word and then insert a scholarly-sounding footnote explaining the entire underlying conceit. The author calls this "a total wimp-out" but then in chapter 9 and 17 he tells the reader "you will meet translators, some famous, who prefer to translate in just that style, and some of whom even try, using pompous scholarly language, to demonstrate the superiority of their wimp-stance. This option is sad." [He's talking about Nabokov, and we're going to see Nabokov treated mercilessly later in the book for his attitudes on translation. And Nabokov has it coming.] The specific problem could "be taken as a metaphor for this entire book, because it is just these kinds of tricky, messy questions that are continually confronted, in varying guises, in the art of translation."
57ff On the various translation challenges of his book Godel, Escher, Bach, with its acrostics and double acrostics, all the puns, the symmetrical and fugal dialogues, the various words and phrases with double, triple, quadruple or even quintuple meanings, the structural puns, etc. The author describes GEB interestingly as "characterized by nearly constant crosstalk between form and content." He has the same types of problems Stanislaw Lem has with his books; and so he went through his book with a red pen to rank all the relative importances of the various verbal games he plays in the text: which were the most important, which could be expendable in the interest of saving others, etc., making a general resource as an aid to as-yet-unknown translators. Also discussion on the unique problems in French because of the author's skill in that language, as well as his hopes that the French version would have the same nuances the English version; and then how he randomly stumbled on the translating team of Robert French and Jacqueline Henry who he worked with on the French language version of GEB.
Poems III: Antique Airs
10aff More versions of Marot's poem in translation; in particular Melanie Mitchell who included all sorts of period-matching Shakespearean expressions, the author thought her version was a "tour de force"; another professor sent him a version pointing out the "semantic chunks" vs the "rhyming chunks" of the original poem which hadn't even been visible to Hofstadter at first and it made the poem in its original language even more elegant than anyone had originally suspected; he shares another version by Melanie Mitchell that embraces this structural element.
12aff The author attempts translations using Shakespearean style, capitalizations and spellings to give it a more quaint, dated flavor; he also offers a "predecessor version" of that same poem--it's interesting to look at the changes he made here; also a metaphysical question that the author asks here about himself as am American cognitive scientist trying to translate a 500-year-old French poem using the style of an English writer whose life began and ended after the French poet even died! "Is this a grotesque, tortured falsification of history? Or is it, quite to the contrary, a reverential bow embodying extreme fidelity to Marot, and thereby rendering a profound service to Truth?"
13aff Finally one of the best translations anyone ever did by William Cavnar, where he managed to maintain the identical rhyme and metric scheme of the original throughout his entire translation, as well as various creative examples of metonymy, as well as certain suggestive turns like using the word "thigh" late in the poem, and rhyming it in the next line with "thy." [This version is genuinely good.]
Chapter 4: The Romantic Vision of Thought as Pattern
63ff On patterns, and abstractions of patterns; the author here talks about a somewhat silly example when he was a child of seeing three balls and deciding that each one was unique but they were unique in different ways: he thought it was both a trivial observation but deep at the same time; he then talks about in the music of Bach there are all these patterns that you can learn and recognize but Bach always fools you, he tricks you in ways you don't expect, so then on the listener/interpretation level you have these interweaved, looping patterns of understanding a pattern/abstracting away the pattern/doing something different from what the pattern says you "should" do, etc. [You see this also in psychology: people have psychological patterns that they play out (see Games People Play by Eric Berne for some great examples), but recognition of the pattern changes the pattern automatically, thus you might move to a different psychological pattern. In Eric Berne's book you see examples where if someone realizes or is called out on a psychological games they are playing, they quickly switch and do something different in reactance. It's another type of abstraction that we do without really realizing it.]
65ff Taking this one level up to the idea of mind and thought: looking at the brain and its thinking activity, but not using the neurological physical level to abstract away where thought happens but at some higher level of abstraction, where symbols and symbol subunits or decoding/coding activity happens; sone of the origins of AI are discussed here, as some of the early artificial intelligence thinkers of the 20th century looked at things in this way.
66ff Discussion of the roots of AI [this will be review for anyone who's read Godel, Escher, Bach]: the French writer Julian Offray de la Mettrie and his 1747 book L'homme Machine, Jacques de Vaucanson (1709-1782) and his intricate almost robotic models of people and animals, and then his "crowning achievement" the automated loom for weaving textiles, from which Pierre Jacquard (1752-1834) then innovated the idea of punch cards that could program the loom with whatever pattern. Then followed Charles Babbage and Herman Hollerith (and what would eventually become IBM) innovating on top of these guys; on the development of formal logic; and then the major setback dealt to formal logic by Kurt Godel; and then the flourishing of study on neurology, information feedback, and what was known as "cybernetics" around World War II; also computer pioneers Alan Turing and John von Neumann and the Turing test.
75ff On the underlying models of AI, not the neurology of the brain per se, but abstracting away higher-level phenomena such as learning, forgetting, foreground/background, categorization, memory organization, analogies and stereotypes, problem-solving strategies, etc.
79ff The author stumbles on a 1964 article in Scientific American by the linguist Victor Yngve on the logic behind the grammars of formal languages, and how it can be captured in a computer program, the author takes this idea and creates his own English sentence generator program. On the authors mixed feelings about the problem of getting computers to use language in a realistic matter: he worried that it was wasteful or a silly thing to do career-wise; he tried math and then physics but then eventually went back to this problem again. [There's an interesting sidebar here on how the author is filled with panic and is kind of crushed to realize that many of the ideas he was toying with some ten years earlier had been discovered independently by others later, and now were well-known techniques in the field of artificial intelligence. It's almost like he was a few feet away from finding gold in the mine, and he stopped digging.]
81ff On machine translation and the intractable problems of the ambiguity of language: on "the pen is in the box"/"the box is in the pen" problem where words can have completely different meanings based on context. Also interesting dueling quotes here on page 82 about the specific problem of translating languages: mathematician Warren Weaver talks decoding something in Russian by thinking that "This is really written in English, but it has been coded in some strange symbols. I will now proceed to decode." [Anyone who has learned even a little Russian would instantly see this as hilariously naive and know instantly that Weaver knows nothing of Russian]. But then the poet and linguist I.A. Richards saying that translating certain Chinese philosophical concepts into English is "the most complex type of event yet produced in the evolution of the cosmos." On the naive simplistic view that language is a one-to-one substitution cipher vs the near-impossibility translation at all. But then on the idea of seeing X as Y, but doing it on an abstracted level, in the form of analogy or metaphor: see the work of Stanislaw Ulam: "What is it that you see when you see?" On the idea of the need to take the word "as" and somehow find a way to mathematically formalize it.
84ff Now onto Bongard problems--again this will be familiar to raiders of GEB. They are fun as heck!
87ff Commentary here on the Eliza Effect: how we attribute "human meaning" to words even if they're spit out by a computer; see Joseph Weizenbaum's book Computer Power and Human Reason; and on the Eliza Effect even traps extremely advanced AI researchers. The author likens it to a child's desire to believe that his stuffed animals are alive.
89ff On Roger Schank and his rejection of syntax in teaching AI's to use language, instead dealing with semantics and a kind of a "script layer" of language meaning: see for example the description of the "restaurant script"; also on the weaknesses apparent here: the restaurant script has huge gaps of knowledge; the underlying idea here is thinking of thought as pattern; kind of an upside down way of looking at language and its meaning; see also so many of Schank's PhD students and their "not always harmonious family of Schankian-style researchers, which call their work CBR "case-based reasoning"; on the idea that people understand situations by seeking analogs in their storehouses of previous experience.
94ff The author discusses certain thinkers who hold towards AI an "antipathy so profound that I could hardly fathom its origins." He cites an essay "Minds, Brains, and Programs" from the journal Behavioral and Brain Sciences. There's a strange cryptic sentence here where the author claims "On account of a conflux of factors sadly out of my control, I must hold off on naming its author, though not for too long." [This seems strange--why be so cutesy about it? But the author is John Searle, and here's a link to the essay]. This paper turned out to be a rallying cry for anti-AI groups; the author points out the (many) structural flaws in the argument: it confuses the human in the black box with a full system of scripts and communications going back and forth with the symbol manipulation that happens; also the scale of this achievement which is duping native speakers with a flawless type of Turing test; also the sheer quantity of text that would be required, potentially trillions of lines of text that would have to fit into this small log cabin/blackbox, and so Searle underplays the tremendous amount of information in play here as well as the human-level understanding of Chinese syntax involved; third he understates the amount of time involved to handle these calculations: the author estimates it would be decades potentially per instruction, in reality Searle is describing a fantastically large and intricate system taking place on a enormous time scale where the human being is basically a drone-like bookkeeper, thus he's not focusing on where the true source of the system's complexity lies. Finally Searle argues that you would be able to communicate completely in Chinese characters but he leaves off many of the linguistic and abstract patterns involved in this translation that wouldn't be possible with a human mind.
Poems IV: Oklahoman
14aff On Xiangsheng, the Chinese verbal art form [you could translate it to English as "crosstalk"], a "semi-improvisatory art" kind of like Abbott and Costello's Who's on First routine; comments on the high quality translation of GEB into Chinese with the help of David Moser; also on Moser's translations of Marot's poem as "Honey Bun": the author describes his rat-a-tat rhythms and rapid-fire colloquial, even slangy, phrases, as well as strict rhyme and meter; see also Moser's version "Sugar Lump" which has certain structural symmetries in terms of punctuation and other aspects; Hofstadter actually flips the poem around and writes it in reverse and it actually works. It turned out that the symmetry in the poem was unintentional and the original author had no idea that the poem could be reversed.
17a Comments here on bringing the original poem into one's own place, time and style vs. leaving in trace elements of its own period and style; also on preserving (or not) its "Frenchness."
Chapter 5: Sparking and Sparkling, Thanks to Constraints
103ff The author asks What actually is translation? Is it the words or is it the ideas? Most people would think of the ideas that need to be preserved from one language to another, thus the words are just vehicles for delivering the ideas--we don't care what the make of the truck is delivering milk just want fresh milk. But if more than one person translates somebody's book the two translations will be different; also we "tamper" to some extent with the work.
105ff [Ah, now I understand why Hofstadter didn't reveal Searle's name earlier on p94-5!] It turns out the author made some "translations" to Searle's article in the prior chapter: here he reproduces the article exactly. What follows here is an awesome discussion of a lipogram: text written with the constraint that one or more letters are not allowed to be used; and now we know we realize that the author rewrote Searle's attack on AI without using the letter "e" [which is kind of shocking: I never noticed, yet it is the most common letter of the English alphabet. This also explains the use of the phrase "Anglo-Saxon" instead of "English"] "I cannot resist remarking that expressing oneself with no 'e''s is definitely expressing oneself with no ease." He mentions entire novels written without the letter "e" and also suggests to certain of his writing friends to try speaking without using any words containing the letter e. "They were keen to try..." The author says it's much like learning a foreign language; then they try to remove both "e" and "t" but then they could hardly speak at all; "...not a single preformed thought could be expressed without lengthy mental struggle, and usually not even then."
108 Excerpt here from the book Gadsby by E.V. Wright, an entire novel written without the letter e. [When you read it, the style is kind of weird, artificial; it reads kind of like a machine translation or a Google translation of another language. But you can't really tell that the "no letter e" constraint is even there unless it's pointed out to you.] The author also quotes the novel's last paragraph, or as the author puts it, its "expiring breaths" [good one! Hilarious faint praise]. Then a description of a novel in French by Georges Perec, La disparition, which also contains no e's--which in French would seem inconceivable. And then amazingly, this work has been "virtuosically" translated into English and German, also without e's! And then it goes further: there's a scene where the characters of the book realize there are no a's in a given passage (and the English translation likewise contains no a's, more on this in a moment) and this is analogized to the book Flatland where the characters two dimensional characters can comprehend one dimension but they can't imagine three dimensions, while we as readers can understand three, two and one dimensions, but not four. And then Perec actually wrote a short story using just one single vowel--the letter e--leaving out all the others; the author does a callback to his day in Warsaw when he was "released" from Polish into French and then into English "each time into a cell allowing me greater freedom" [see page 49 above]. This gets the author to think if there might be some master language or meta-language that was like a "fourth dimension" that would allow him to express himself better even than he could in English.
115ff On the passage in the novel without e's that also had a passage lacking both e's and a's: in French Perec bounces between tenses in order to solve for the constraints [there's a discussion here for French geeks about how the author bounces between present tense and the passé simple (French's literary past tense) to solve for the no e's and no a's constraint here]; but then Hofstadter goes over a simple three word challenge "un loup fuit" to be translated into English ("a wolf flees/runs"or "a wolf fled/ran") where that translator solved it by translating into "chipmunks run wild" because the English version also can't contain any letter e's or a's in this passage. Also the German translation of this, which also lacks the letter e, uses this passage in a translation that lacks the letter r.
117 "In the act of translation, there are always two 'frames'--the culture of origin, and that of destination--that inevitably get blended in countless ways as the ideas are transplanted from one soil to the other. Some ideas transplant easily, others put up a fight, occasionally ferocious, and some simply will not go at all, no matter how hard they are shoved."
118ff Now we get into the meat of the idea here, which is the battle between content and form in a translation. Note that machine translation focuses on content to the exclusion of form. So in a machine translation there's no asking what about when form and content are intertwined, it is assumed that they're independent and always independent. The language "level" is as superficial as say the font of the text. The problem is that words are structural in breaking the world up into form, see for example the word "Uncle" which means different things in different languages [or see an example of the subjunctive or the imperfect past in a romance language versus English, or the passive verb forms in Spanish that don't exist in English in the same structure]. Note that certain types of contexts--like a technical manual for a washing machine, or a weather report--where there are specialized terms and clear demarcations between medium and content, in these micro-worlds, yes, you might be able to "pour" one language into another and vice versa, but this absolutely doesn't work when there is a mingling of form and content in say a poetic work, and certainly not in an example of a novel lacking a certain letter.
124ff Now we take things to a whole new level, to Guiseppe Varoldo's Italian "translations" of Western literature classics into synopses, each perfectly constructed into Italian-style sonnet that only use just one vowel. [This stuff is unbelievably crazy, it's absolutely inconceivable to do this in Italian.] See for example this first stanza of a sonnet synopsizing Dante's Inferno:
Nel mentre ch’è trentenne, l’Eccellente
(nelle Lettere regge, è legge, splende)
ben nel ventre terrestre se ne scende:
ente perenne, sede del Fetente.
128ff On the idea of an "esthetics-based transfer of essence from one medium to another." On ShrinkLits, Maurice Sagoff's book distilling 50 canonized works of world literature down into distilled poems.
132ff Interesting point here from the author: a big part of this process starts with the author selecting only what works will "work" within the constraints he's using, so the constraints aren't quite as restricted as they appear. The author knows what to avoid. Also comments here on Tom Lehrer and his work, the author had lunch with him in 1984 and got him to talk about his work and how he dealt with "the self-imposed constraint of rhyming" and Lehrer said it wasn't really creativity it was just the constraint itself which caused him to dream up bizarre images in order to make rhymes happen: it was the rhyming constraint itself that made him look more ingenious than he was; see the example here of rhyming "industrial waste" with "toothpaste," two images that would never come together unless there was a rhyming constraint. The author calls this a "free assist." Then there's a game the author played with his friends in college, the "left-right ride" where they would take a zigzag route, alternating left and right using a simple rule, and it would cause them to find unbelievably obscure nooks and crannies in their town. It's an analogous constraint just like rhyme that brings about a process of discovery, in this case causing them to visit places that they never would have seen otherwise.
135 Rhyming with the (difficult to rhyme) word "orange" by breaking it up into two phonemes. This example is one of many the author shares:
Eating an orange
while making love
Makes for bizarre enj-
oyment thereof.
136ff On The Raven, as well as Edgar Allan Poe's essay "The Philosophy of Composition" where he demonstrates the creation of the poem and how it was inevitable from the very start, he likens it to a math problem. Comments here on "translating" that poem into English without the letter e, using the self-referential nom de plume of Arthur Gordon Pym, a character from Poe's own novel The Narrative of Arthur Gordon Pym.
138ff "Robert Frost once made a delicious remark to the effect that writing free verse is like playing tennis with the net down." Also criticizing and disagreeing with Frost's comment that "Poetry is what disappears in translation." The author retorts, "this whole book is a riposte" to that idea.
Poems V: Sue Suite
18aff More versions of Marot's "Ma Mignonne" by David Moser: using constraints of two-syllable lines; then Hofstadter's own version "repairing" it by adding back a syllable, but making two of each of the syllables in each three-syllable line internally rhyme [this poem translation lacks character and it's a little try-hard but I get the underlying idea of playing with constraints] but then as this version was too long he starts pruning it down, and then discovers that the version that comes from this process "had given rise to a new type of fidelity to Marot."
Chapter 6: The Subtle Art of Transculturation
141ff On the creation of the Chinese version of GEB, and the author's nervousness because of how the Spanish and German versions were terrible and ignored all of his translation notes. Note the hilarious offhand quip about how Peking University is "oddly retro-yclept."
143ff Interesting comments here on palindromes, how in English they exist on a letter level but in Chinese they exist more on a word/morpheme level. "A man, a plan, a canal, Panama" does not look like a palindrome to Chinese-thinking person reading in English.
144ff On the ideal that one should strive for in translation: faithfulness, clarity, and grace.
145ff On the process involved in writing the Crab Canon in the GEB, a dialogue that was palindromic on a word level, line by line; he started by collecting phrases in English that had two utterly unrelated meanings like "not at all," "one has no frets," etc. Thus literally translating it into another language would have no value at all--it would be incoherent on every level, because the form matters more than the actual content .See also other examples of crab canon-style works, like Bach's own musical Crab Canon, M.C. Escher's drawing of crabs moving in different directions; also a discussion of DNA that has aspects of Crab canon-ness.
148 [This chapter has had two different examples of an interesting form of what I might call "translator solipsism": first on page 144-145 Professor Wu is talking about a translation of the Diamond Sutra from Sanskrit into Chinese where one version changes the Indian placenames into cities that would be known to Chinese readers, rather than the original placenames (which would not be any place Chinese readers would have ever heard of). The Chinese professor was bugged by this, whereas the English speaking people he was discussing the idea with thought that it would actually be helpful because it would give someone from another frame of reference context that he wouldn't otherwise have, rather than context that they could never have; then see also on page 148: there's another really interesting example here of figuring out how to translate the phrase "Speak of the devil!" into Chinese. The original literal translation came back as "Mention the demon, then comes the demon!" But at the same time there is actually an idiomatic expression in Chinese, from the Chinese historical novel Romance of the Three Kingdoms that would have worked, but the Chinese translators didn't want to use that phrase because it was too "steeped in Chinese culture" such that only a Chinese author would have thought of it: thus if they used it in the translation it would sound like the book wasn't written by Hofstadter! There would be no way in American would know about this expression. But then American translators were saying basically, Wait: Chinese readers will obviously know that Hofstadter didn't write the book directly in Chinese, they will know it was translated by a Chinese speaker: Hoftstadter writes, "Their [the Chinese translators] protestations were vehement: 'If we adopt your strategy for rebuilding the book in a pure Chinese style, then readers will feel the book is no longer by Hofstadter.' And yet from my perspective, it was precisely the reverse--I would feel the book was no longer by Hofstadter if the team failed to adopt the translation strategy advocated by Hofstadter! Could it be that the very idea of transculturation itself is a Western one, and strikes the Oriental mind as alien? In that case, paradoxically, the most deeply Chinese-style translation of GEB would be an ultraconservative one in which all Chinese idioms and metaphors were carefully eschewed, in which all Chinese vignettes or situations were strictly shunned, in which all references to the Chinese culture were religiously avoided--in short, a translation in which the book's rootedness in America was as blunt and stark as a good slap in the face. Conversely, and equally paradoxically, a translation of GEB such as I myself was so keen on--keen to the point of having sent, at my own expense, a personal emissary over to China to help realize it--in which every English-language pun and structural game was ingeniously re-created in flawless, sparklingly witty Chinese, in which all references to American geography, history, or culture were seamlessly replaced by lovely and crystal-clear Chinese analogues, would strike Chinese readers as being a weird, trumped-up, artificial, 'foreign-style' translation. By virtue of being overly Oriental, it would be extraordinarily disorienting!" [This is fascinating on a few different levels]
149ff Note the author describes here how he deeply upset he was; he saw their attitude as perverse and inscrutable, but then he thought of Solzhenitsyn's novel One Day in the Life of Ivan Denisovich: it's about Russian prisoners, and the author read it in a slangy American English translation, which felt totally wrong to him, the novel had been over-translated. He wanted to get the vibe of that period, not see a modern American version of it, where the translator used "guys" for zeks, "mush" for kasha, etc. He understands now that the Chinese view on the translation was actually a type of "faithfulness" that he himself "had embraced in a different circumstance" but he still disagreed with their approach in the instance above: he still thinks that culture is part of the vehicle to convey the underlying image, although there are parts of his book that are culture-dependent and some that are culture-independent.
153ff Discussion here on "transtemporation" (e.g., taking Little Women and translating it into modern Italian, or translating Shakespeare into another language but using modern speech patterns) vs "retrograde transtemporation" (taking JD Salinger and translating his work into Shakespeare-era English, basically the opposite of translating Chaucer into Modern English). See also comments here on Shakespeare translated into German in the 19th century by August Wilhelm von Schlegel and Dorothea Tieck, which were considered so amazingly well done that certain critics said that Shakespeare actually became German [!] and even that it was "written inside the accidental husk" of English. [This entire discussion is really interesting, very subtle, there doesn't seem to be a consistent set of rules for what is appropriate, or for what works well and what doesn't work. Note also that the Louisa May Alcott Little Women example completely contradicts the author's experience above with the Solzhenitsyn novel! The one "worked" when modernized and acculturated, the other one absolutely didn't.]
155ff Comments on Shakespeare, how it can be rendered in other languages and modernized; but to a Shakespeare purist, "modernizing" it into modern English somehow seems gross. Note also that Shakespeare's plays are unbelievably flexible and can be fit into different domains; they were even set in various domains by Shakespeare himself: see Romeo and Juliet which takes place in Verona, Italy, then also see the musical West Side Story which is the same story, but rendered in New York city.
156ff Examples of trans-temporation of the ideas behind words: Hofstadter cites a "peculiar" example of the original foundation documents of Stanford University in which founder Jane Stanford specified no alcohol to be served on campus or within three mile radius of the university; this was later "reinterpreted away" using the logic that had Stanford been alive today she would have the social mores of today, not of the Prohibition era [this is hilarious: by this kind of "logic" you can reinterpret anything to fit current fashion...]; we also reinterpret literal Bible lines in order to accommodate a heliocentric universe or evolution, even though those notions were once heretical. [I'm not sure quite if this is the same kind of problem: there's a difference between the extreme stretch of "reinterpreting away" a clearly stated prohibition like "no booze" and applying timeless biblical wisdom to periods centuries later. It's not like we are "reiterpreting away" Thou shalt not kill."] Also comments here on old music with low fidelity sharpened and cleaned up: see for example Louis Armstrong's St. James Infirmary, see also movies that are cleaned up or colorized, etc.
158ff On dubbing versus subtitles in movies: funny anecdote here someone in Italy who told the author's wife that she adored the voice of some American movie star, but she had actually only heard the Italian dub artist! See the Groucho Marx movies that are redone in Italian where Groucho's witty comments are done in Italian and are utterly different from the English original, such that Groucho Marx has "a second identity in Italy." Also [I never knew this] Laurel and Hardy translated into Italian (as well as French) uses voice actors who have very strong American accents, this adds to their silliness, somehow it adds to the zaniness of it.
159ff Lengthy discussion here of the 1987 book Chinese Lives: An Oral History of Contemporary China which features interviews with workers in China in the style of a Studs Turkel book; it is translated into modern-sounding English with expressions like "ain't got no" and "a load of crap" and words like "hick" and "softies." The author parodies it by running his own translation including expressions like "no tickee no washee" and "not a Chinaman's chance," etc. Obviously, the author wouldn't use crude English stereotypes of Chinese people, but the idea is to show that it's not really genuine if spoken in casual English language--this strips the story of its Chineseness. The author calls it "ludicrous and grotesque."
165ff Another interesting example from GEB, describing the author at an American football game: the French version was translated where the author was with another man (in the original version the gender of the person he was with wasn't specified) watching a rugby game, because French people would understand rugby much better than American football; the author didn't want this version for two reasons: first, he had never actually been to a rugby game and considered it a distortion that was not faithful to the story, and second, he was with a dear friend who happened to be a woman so it was an affectionate memory of his. And then the translators collectively did pretty much the exact opposite thing when they translated the descriptions of how he built his Crab Canon, saying that the author had been collecting French (not English) expressions with double meanings, which is an even less true version of the reality! But they collectively decided to leave this in, even though it was overtly false, and even though French readers--if they really thought about it--would know it was false because GEB was written in English! The author explains that the ultimate solution here seemed to have more simplicity and elegance, overwhelming need to tell the literal, exact truth. [This is an extremely subtle problem that you would only bump into if you were lucky enough to publish a book that saw enough demand to be translated! But the problems are interesting to think about.] Likewise in the Chinese version of the Crab Canon they worked with a falsified version of the author's experiences, coming up with various double meaning expressions in Chinese which of course could never possibly happen with the American author Douglas Hofstadter, but it makes the story work better.
Poems VI: Bold Ventures
22aff On Robert French's "one syllable per line" translation of Marot's poem, with beautiful form; also a version of the poem by Melanie Mitchell, "My Minion," that uses French-English cognates, this version is weird but extremely interesting, although it only makes sense if you speak some French.
24aff A version by O.B. Hardison that the author tampers with and totally revamps to fix some of the glitches that bugged him (the original was just improvised, it's a very terse version), and then another brief version by the translator Michael Kandel, translator of Stanislaw Lem's science fiction works; and then a discussion about how the author wrote to Kandel to encourage him to rewrite and translate Lem's sexist pros in a less sexist way. Kandel ultimately made certain changes but also replied that Lem's "sexism doesn't offend me. It's a weakness, and not his only weakness. The wisest and most enlightened of us have their weak, foolish sides." Finally Hofstadter's own mother does a translation, and he makes some interesting comments about his mother's style, he owes his love for words and language to her, but her verbal style "tends toward anarchy, irregularity, and capricious unpredictability." He's actually disappointed and kind of mad at her for her version, but then later came around to appreciate it, in fact on some level he thinks it's the most faithful to the spirit of the original, despite not having any of its formal properties.
Chapter 7: The Nimble Medium-Hopping of Evanescent Essences
171ff On translating chess from context to context like from a square board to a triangular board, or to a hexagonal board, as a metaphor for changing Marot's poem from the square-based Anglo-American "board" to a triangle-based German "board," etc. The author interjects here a nonsense poem and all sorts of weird words that don't really make sense in English, and then a discussion of what would be a Bishop's move in chess on a hexagonal board? There's no such thing as a "diagonal" here--it doesn't translate, so the author thinks about different ways to think about diagonality on a hexagon-based board; on the idea that "diagonality" would have to become a metaphor in order to be lifted out of a square chess board and plunked down into a new medium: will would have to let go of its "surface attributes" to get at its essence. [This sounds quite a lot like the author's discussion in GEB of "lifting and skimming off" intelligence out of the "hardware/software" matrix of the human brain to try and translate it somehow into AI/machine intelligence.] The idea here is "the transport of an elusive essence between frameworks" here literally rendered in a metaphor of movements on a chessboard.
174 The author sneaks in a quick poetry break, a poem filled with strange, slangy English words.
178ff Comments here on "grain size" which depends on the relative sizes of the underlying structure versus the top level structure; this is yet another metaphor for faithfulness in translation; the semantics and syntax differences between Russian and English will give tens of thousands of metaphorical "jaggies" (or deviations) from the original shape, but these are fine-grained, word-based and sentence-based deviations that should even out over the larger "grain size" of a book. The problem with a short poem or a single sentence is that the grain size on the word and syntax level is much larger relative to the work. That's why coming up with the right (translated) title for a book, or a good translation of a poem takes a very long time: "you can't shove the grain size of human language under the rug, because you are at the grain size."
182 Cute sentence here in "Anglo-Saxon" (the author's name for "e-less" English): "As an illustration of this notion, Anglo-Saxon will do a good job, for in it you can say most anything you wish (as in fact this thirty-word auto-alluding affirmation shows)."
182ff On how a linguistic medium is not just at the language level: on monosyllabic English as a linguistic medium, or other restrictions on length and style: say for example the "medium" of a sonnet; but the author doesn't think of "palindromic sentences" as a linguistic medium because you cannot say anything you want in it, in fact you can say very little in it. :ikewise the set of words beginning in x cannot be a medium.
183ff Another poetry break here, this one a French poem by Francois Villon, who is well educated but wrote poetry using "medium" of the jargon of Paris riff-raff. It turns out that even if you speak French well, you'll need this poem to be "translated" into understandable French. This is the same exact situation with the slangy English poem on page 174 above, which likewise needs to be translated into understandable English, because it makes no sense to a modern speaker. It turns out that the slangy English version is a translation of Villon's poem from guttural French into equally guttural English, done by the poet William Ernest Henley, author of the famous poem Invictus [this is the "I am the captain of my soul" poem.]
186ff Biography of Francois Villon, along with more of his poems, with lots of double meanings and use of jargon of his day that "eludes virtually all French speakers today, even specialists."
188ff On Henley's translation of Vallon's poem, using 19th century (or older) British slang, much of it British underworld slang, to "fig a nag" or "knap a yack," etc. The author spends the next few pages explaining how he tried his best to figure out what the various expressions meant by context, and by extension how to read/translate the poem. "It's a strange thing to find yourself entirely at sea in your own native language." But yet it's still a "linguistic medium" in the sense that the author talks about at the beginning of this chapter.
192ff Discussion of certain unarticulated components of a translation that may not be concrete or front of mind for the translator, but yet still exist in the poem: the author gives an example of the "Cutie Pie" version of Marot's "Ma Mignonne" that included "flush your phlegm," or the "Hi Toots!" version his mother wrote; these show the jaunty, flippant tone of the original poem, but in ways that are not literally in the medium of the original, thus it's a creative liberty taken in a translation that happens at a different level; there are two very different levels in this creative act here: solutions found within the medium and the medium itself that's chosen by the translator. "...a word or a phrase in a given translation may at first be perceived as a breach of good taste and rejected out of hand, but then may, with further consideration, come to be reperceived as fitting well within the norms established by a linguistic medium that one at first simply had not recognized as such, or perhaps that one had recognized but rebelled against."
195ff On the song "The Surrey with the Fringe on Top" from Oklahoma!: does it make sense to even translate this at all? But then there's a version of it in German performed by Marlene Dietrich, but it's been "transmogrified": it's a horse-drawn sleigh out in the snow, it's winter time, the woman is singing (not the man as in the original) [note also the highly unfortunate title of this song: "Schlittenfahrt"--yet another example of a translation problem!]; the author also describes a German language country western music radio station that he stumbled onto one time where everything was sung in German, but yet somehow with some kind of "universal country accent"; see also My Fair Lady translated into Spanish and into Hungarian; the Hungarian version is fully in Hungarian, but there sort of the charade maintained that she actually is speaking English; the author talks about the "weirdness" of these versions; also the author makes the rather appropriate observation that, yes, it is a suspension of disbelief to hear dialog in Hungarian where everyone on stage is pretending it is in English; but this is no more a suspension of disbelief than the one where we have to accept a story punctuated by actors breaking into song on stage and then returning back to speaking after the song is over.
201ff The reader gets a lecture here on non-sexist language and the various latent sexism structurally built into English; we even get a multi-page lecture on the specific sin of using the phrase "you guys" [which happens to be one of my favorite catch-all terms of address! Unfortunately the author sounds like an absolute Boomer here decrying the "deplorable state of affairs" in "linguistic self-awareness in our country today." As a millennial today might respond: "Thanks Dad!" Note that Douglas Hofstadter was born in 1945 so technically he is a Silent, but barely so.]
204ff On linguistic isomorphisms to "you guys": see Holland and the word jongen; similar isomorphisms in Chinese, a word that translates into older brothers [the book seems to be losing its way a little bit in this chapter, I think a stricter editor would have demanded he cut out about half of this chapter.]
207ff On Mary Daly and her Wickedary of the English Language, basically a lesbian/anti-male slang dictionary; the author uses it to write a poem that a version of H. De Vere Stacpoole's 1914 translation of Villon's slangy Ballade; thus we get a third "poetry break" here, called "D's Cross Tip to All Straight Wives."
212ff Now the author discusses jokes and joke telling: we get lectured more here on how they reveal pernicious sexist attitudes; and then section on "joke deporning": removing the sexual element in a joke and seeing if the joke remained funny: can you translate a sexually-explicit joke into a non-sexual setting and at the same time translate its funniness? [Safe to say the answer here is a clear no.]
216ff On "slippage humor": humor based on the sudden unexpected slippage of some mental structure that seemed solid until it didn't: see Mel Brooks' Silent Movie where the one spoken word in the film is said by Marcel Marceau, the famous mime].
220 On the ur-joke: when a joke has an underlying structure that can be transferred successfully from one medium to another; where two jokes share a structure so that they're "the same joke"; note that the ur-joke is not funny in the least and takes longer to explain than the actual joke itself.
224 On the French mathematician Raymond Queneau and his unusual book Exercises de style, basically telling the same banal story, but from 99 different perspectives and linguistic mediums: a type of verbal variations on a theme like Bach's Goldberg Variations. The author takes care to distinguish this book's lack of artistic depth (it was a literary experiment) compared to Bach's "monumental construction." Comments here on the translation challenges involved in bringing Queneau's book to English; some of the versions of this completely unremarkable story range from telling it in transcribed foreign accents, in regional accents, in a version without the letter e, in slang and colloquial treatments, variations that play with tense, mood, point of view, via metaphor, retellings versions that are legalistic, poetic, scientific, obscure, terse, verbose, in sonnet form, in playlet form etc.
228ff The author describes it as an ur-message, and also notes that it has been translated into several languages, including, notably, a translation into Italian by Umberto Eco. "How would one go about translating a book that is not, like most works of fiction, a series of messages in a single medium, but rather, a series of media supporting a single ur-message?" On becoming acquainted with a Dutch translator of this book who wrote in his introduction that since the book consists of diverse media, translating this book requires jumping "to a higher and more abstract level of translation than is usual" because you have to translate linguistic media from one language and culture into a second language and culture, like placenames, jargon, accents, etc. [This is yet another section the author could boil down considerably.]
231 Taking the metaphor of the ur-theme or the ur-story/ur-text in Queneau's book, and thinking about the ur-dog: what the variations of all the dogs out there--all their breeds--what are they variations of? We can't really pin that down: the more dogs you see the more aspects of the category "dog" you can know, but then "your image of the ur-dog gets continually blurrier." The better you know it the less you can put your finger on it, and the same elusiveness applies in other domains (there will be more discussion of this in Chapter 11 as we look at categories). Finally the chapter closes with Hofstadter's own version of the ur-story using a chess example, calling back to his "hexagonal chess" example from the beginning of the chapter. Interesting.
Poems VII: A Gala of Gists
27aff The author makes a fortuitous discovery that the phonemes of the name "Clément Marot" can be rearranged into the French words "Mots-clés marrants" ["funny keywords"], and so he builds a poem of English keywords that illustrate the [French] poem's context:
Entrée.
Pleasantries.
Commiserations.
Encouragement.
Poietautonym.
Enticement.
Concerns.
Pieties.
Echo.
[Note also the acrostic palindrome, EPCEPECPE, that the author calls a symmetric "keychain."]
And then in English this version:
Poet greets dear.
Poet sheds tear.
Poet hints cheer.
...etc.
30aff And then more versions that combine and rework these original versions; and then a 10-line meta-poem that maps all the formal properties of "Ma Mignonne." All of these various meta-poems are experiments influenced by the theme of the book Exercices de style by Raymond Queneau.
Chapter 8: A Novel in Verse
233ff [I'm grateful to the author for this chapter because it finally got me to read--in translation--Eugene Onegin.] The author describes his very weak Russian; he mocks himself here for having the idea that he would actually write about or even judge various English translations of Pushkin's poem Eugene Onegin; not only is his Russian very basic and weak, but he hardly knew anything about Pushkin.
234 Cute quote here: "...the nice thing about having a brain is that one can learn, that ignorance can be supplanted by knowledge, and that small bits of knowledge can gradually pile up into substantial heaps."
234ff He stumbles onto a novel in verse called The Golden Gate from the mid-1980s by Vikram Seth which was a literary incarnation/variation of Eugene Onegin; also a reference here to Charles Johnston, the British poet, who wrote a highly regarded translation of Pushkin's work in 1977, followed by yet another translation by James Falen, likewise a very high quality; Hofstadter and his wife decide to read the two translations together, trading the books back and forth, reading corresponding stanzas from each book. [What a beautiful thing to do here!] "For me, and I think for Carol as well, the intense sharing of these two translations was one of the high points of our all-too-short married life, something I am so grateful for having done."
237ff On the "Onegin stanza": 14-line stanzas written in iambic tetrameter, rhyming ABAB, CCDD, EFFEGG. Also on a fixed distribution of masculine and feminine rhymes (e.g.: plagiarize/evade your eyes). Note that both Johnston's and Falen's translations of Pushkin's poem retained this strict structural form.
239 Good analogy and callback here where the author talks about his exposure to Bachs' Well-Tempered Clavier, preferring it on the piano rather than the original harpsichord; see also Ravel's orchestration of Modest Moussorgsky's piano work Pictures at an Exhibition; and then thinking back to the analogous thesis by Gundolf that Shakespeare had reached his "manifest destiny" when he was translated into German [see p153-4 above]. On a work of music or a literature becoming an abstraction that floats above its original medium. Also on the question of what, actually, are you experiencing when you're reading a translation of Pushkin? You're not experiencing Pushkin directly, but aren't you still experiencing Pushkin? Otherwise what else were you doing? Hofstadter also notes how he'd find some great turn of phrase in one of the two English translations, and he would catch himself saying "What a fine touch Pushkin had!" How even do publishers get away with writing "Eugene Onegin by Alexander Pushkin" on the cover of books in English?
243ff Now for a very interesting four-part comparison of four different translations of four different stanzas of Eugene Onegin. "These anglicizations are a bit like performances of Chopin's Études in the media of string quartet, woodwind quartet, brass quartet, and barbershop quartet--each one faithful and unfaithful in its own way to the original."
246ff The author's favorite is Falen's translation, "an unbroken string of beautifully chiseled Onegin stanzas in our language, as rhyming and rhythmic as those of the Russian original." "But how would I know? I've never read the Russian original!" Discussion here on the slippery performer/composer conflation we have in music or the idea (again calling back) of an Italian friend who thought she "knew" the voice of her favorite American movie star but only knew it in dubbed Italian!
249ff On triangulating back to the original text, looking at the different translations and seeing where liberties were taken that differ from the other translations, as well as where the versions agree; then you can make inferences on how the original Russian probably was rendered.
251ff On the idea that Eugene Onegin in Russian is a lot like Shakespeare or the Bible in English, where so many of the phrases from this work have entered the language that they sound familiar, like a type of déjà vu. Also comments here on the ordering of reality and fiction in this work, as Pushkin himself claims to have the letter that his friend Yevgeny got from Tatyana, although this still doesn't explain how the author would have an intimate treasure letter written by his fictional friend. Also there's another weird parallel with reality in that Pushkin himself was killed in a senseless duel, and in this poem the other principal male character of the novel, Vladimir Lensky, is also killed in a senseless duel, shot by Eugene Onegin himself. [!] There are other examples of self-referential twists in the work, like the letter that Tatyana wrote in artless country girl French, the author Pushkin admits to his readers in the poem that he translated it "poorly" into Russian, but yet it is written in beautiful, rhyming Russian tetrameter, perfectly fitting the Onegin stanza structure. [Once you read Eugene Onegin, yes, these self-references sure do bend your brain...] Also what if you translated this poem into French--which has been done--will the letter, originally written in French, now sound more genuine because it's in its original linguistic medium? [!!!]
Poems VIII: Sassy City
31aff The author uses enjambment [the continuation of a sentence or an idea from one line or couplet of a poem to the next] as a "feature rather than a bug" in one translation of "Ma Mignonne." Here's an excerpt:
Your sweet tooth
Has been ruth-
lessly squelched.
32aff Other versions here that use slangy English, another using homonyms, etc.
34aff The author gives "translations" of "Ma Mignonne" riffing off the George and Ira Gershwin song "But Not for Me"; see Hofstadter's version here "Hurry, Love" where he starts to get even a little bit cocky with his language manipulations: "Quarantine's / A mean means" for example; also a cute story about how he and his wife loved 1930s and 40s-era songs by Gershwin, Cole Porter, etc., basically American popular music from the era before they were born; he tells a story of how they met and went up and spoke to Frankie Gershwin (George and Ira's little sister), who happened to be sitting in the same theater watching a Broadway show. His wife described it as "like shaking hands with an era." Finally the remaining two versions of the poem take various risks with rhyme, enjambments, adding in subjects, homonyms etc. The author asks are these various versions properly translations? Or are they "interpretations of" or "poems inspired by"?
Chapter 9: A Vile Non-verse
255ff Referring back to George's Perec's "e-less" novel, and pushing that idea further, asking if a translation of this novel without the e's would be necessary [i.e., making sure to maintain a key form/structural element], or would such a thing be too slavish to form and thus somehow be a betrayal of literary truth? The author now posits a hypothetical playwright "Marvin Validbook" [which looks like an anagram for Vladimir Nabokov, it turns out it is, and the author is satirizing--actually making a reductio ad absurdum argument here against what we will come to find out is Nabokov's very rigid philosophy of translation]; the author also is getting pretty preposterous here with puns and alliteration here, it's kind of funny.
257 Now on to discussing Vladimir Nabokov directly, who translated Eugene Onegin into English literally word for word without any metrical or rhyming aspects of the original; his view was it would be a betrayal of literary truth to imitate these structural properties in English. Nabokov calls rhyming Pushkin or translating him with this form "mathematically impossible," oddly. We come to find out our initial character Marvin Validbook writes everything that Nabokov writes but without the letter e. We find that both these authors are guilty of the same types of errors--and the same type of hypocrisy. Note that Nabokov harshly criticized the scholar Walter Arndt, who did a metrical and rhyming translation of Eugene Onegin that won an award; Naboko criticized it for being a "dainty mimic." This and some of the other stories here about Nabokov are quite hard to believe; the author talks about how Vladimir Nabokov had "a strange sinister streak in him."
259 Quoting Nabokov from the Foreward of his translation of Eugene Onegin: "To my idea of literalism I sacrificed everything (elegance, euphony, clarity, good taste, modern usage, and even grammar) that the dainty mimic prizes higher than truth." The author responds that this view would be okay "were he not insufferably smug about it."
260 Comments on Nabokov's commentaries along with his translation of Eugene Onegin, some 1000 pages in two volumes that the author considers astonishingly erudite: all sorts of commentary on Pushkin's poetry, biographical information, context on Russian and European history, etc., the author likens it to Borges' character Fune in Funes the Memorious, and also calls this an unparalleled sourcebook, and that Nabokov "did a great service to Pushkin and indeed to Russian literature."
261 Hofstadter quotes some of Nabokov's extensive [really, unending] commentary on various French poets that Nabokov liked because they were precise, rather than "shackling themselves with trivial and treacherous rhyme." Also there's an interesting section here where Nabokov refers to some lines by Marot, along with a hilarious comment from Hofstadter: "Well, Well--of all people to turn up at Nabokov's big Onegin bash!" [Separately, it's quite an interesting experience to read these lines work in 16th century-style French: it uses older spellings such that you can really see the Latinate origin of French here, it looks like a hybid of Spanish, Italian and French all together. See photo below:]
"De soy mesme" [De soi même "About Myself"] by Clément Marot
262 A couple of things here worth noting on this page: first of all Hofstadter subtitles this section "A Mean Wordsmith Wielding a Mean Wordsword" and then goes on to describe Nabokov's harsh criticisms of other translators of Pushkin's work describing them variously as "unfortunately available" and "in burlesque English" and "grotesque satellites." [I think I might have to file away "grotesque satellites," that's a good one.]
263ff Next Hofstadter offers four stanzas from Nabokov's own translation, the very same four stanzas that we compared in Chapter 8 [p243ff] from four different English translations of Eugene Onegin. And quite honestly, Nabokov's "translation" is literally unreadable. I'm sure it's all perfectly accurate and everything but it is awful. Everything is lost without the form and structure--most of all what's lost is any shred of readability! Hofstadter offers his own faint praise: "Perhaps it would be most appreciated by students of Russian taking a literature course in the original language." One other use [which I'll add was cited specifically by Charles Johnston in his translation] it is an exhaustive guide to find out precisely how Pushkin actually "said" his poem in Russian.
266ff Comments on the internal inconsistency in Nabakov writing about Onegin because he also took pleasure and pride in translating Russian poetry into English and vice versa; in fact, his first book was the Russian translation of Alice in Wonderland rendering rhyme into rhyme. Even in his commentary on Eugene Onegin we see him translating poetry into poetry (with rhyme) just as he criticizes these other writers. Also comments here on the fact that Walter Arndt's translation, in verse, of Eugene Onegin came out at the same time as Nabokov's multi-volume commentaries came out. ["Baaad timing!" as Hofstadter puts it.] Nabokov savages him, calling it a "monstrous undertaking" and describing him as a "pitiless and irresponsible paraphrast" seeing the work has "more or less sustained stretches of lulling poetastry." Note this other good Nabokov insult, referring to a "chancrous metaphor" in Arndt's work. Hofstadter comments here about how Nabokov assassinates other authors using "words that I never even knew existed!"
269 Interesting section here on Nabokov being a genuine cunt to his own former friend, Edmund Wilson, who rose to the defense of Arndt's Onegin translation.
270ff Walter Arndt, unfazed, actually came out with another anthology of Pushkin verse in 1972, Pushkin Threefold, with both commentary and translations of lots of Pushkin's work, including a response, direct and indirect, to the insults from Nabokov. "Nabokov's recent two-volume commentary in English on Eugene Onegin attempts to call into question again, not a verse translation, but verse translation itself. This happens once or twice in every literary period." Interesting discussion here on Arndt's point: you have to transmit the "import" as well as "impact" both cognitive as well as aesthetic of a poem, not render it in garbled prose while assuring readers that "the corpse in its lifetime was poetry." [Ouch] he calls Nabokov's approach "ritual murder" followed by "lexical necrophilia" [Double ouch]
271 Also more of Arndt's discussion here of "what is the unit of fidelity" in a translation, and how does one maintain simultaneous "double fidelity" to both content and form.
272 Comments here from James Falen on his translation of Eugene Onegin where he attempts "to seek solutions without self-indulgence" and "to earn a freedom within the bondage of the form"; in other words the rigidity of the stanzas of the poem become like building blocks, a structure in which the translation is formed. Also in an exchange of letters that Hofstadter had with Falen, Falen included a six line poem, that Hofstadter says "could be taken as the theme song of this book."
Every task involves constraint,
Solve the thing without complaint;
There are magic links and chains
Forged to loose our rigid brains.
Structures, structures, though they bind,
Strangely liberate the mind.
274ff Note another hilarious irony on Vladimir Nabokov's book: it doesn't even use Russian Cyrillic text! It uses "transliterated" Russian in English characters. [This is hilarious! What a loser.]
276 Note the quote "incapacitating ictus" [fatal/severe stroke] and various other prosodic Easter eggs here used in a passage about Diana Lewis Burgin, a professor of Russian literature, who wrote a biography of their father in Onegin-style stanzas in English. "The day before her father suffered an incapacitating ictus at age 88, she'd bought a tape recorder to tape his memoirs, but all her plans went up in stroke and so sadly she madly scrambled to try to reconstruct his life."
277-8 Final comments on how condescending--and self-revelatory--Nabokov's comments were about various translations of Eugene Onegin, the author here describes Nabokov's translation as tedious, awkward, sparkless, and "the diametric opposite of reading the original."
Poems IX: Two Little Families
37aff Here the author offers a fairly literal translation of Marot's "Ma Mignonne' called "My Petite" and then plays with this over the next few poems, ultimately arriving at a gemstone conceit that he never expected.
38a Note this comment on the second version: "...one becomes quite unaware of the degree to which one is forcing and stretching the language to one's own purposes. For this reason, I have often been rudely shocked, when listening to someone else read a poem of mine out loud, to hear my carefully constructed lines accented wrongly left and right, to hear my precious little couplets brutally mangled, when I naively had thought there was only one conceivable way of reading my creation out loud. Such shocks have taught me how hard it would be to come up with a poem that virtually any native speaker would effortlessly read aloud with the timing and rhyming of every line coming out just as the poet hears them internally."
39a Another cute version of this poem that has both "Doug" and "Clem" [the translator and the original poet] as characters; and the middle line (Doug/Clem) is to be read Doug "slash" Clem. / So smash flu.[!] Also on the author's obsession with working the word "palpitate" into this poem, and it caused him to work it out backwards because he knew that the next word would be "heart."
40aff Comments here on his notes and working word combinations for other translation versions; also a series of versions based on one that he tossed off in half an hour called "Little Gem"; this is an interesting case study in the "genuine unpredictability of the creative process" as he next discovered there could be a conceit here about a gem, and he thus worked on this conceit over the next series of variations, finally arriving at the "unanticipated destination" of "Gentle Gem," his 15th [!] version of this set of revisions.
Chapter 10: On Words and Their Magical Halos
279ff On Pierre Menard, the novelist who was a figment of the imagination of Jorge Luis Borges; Menard, per Borges, wrote the unfinished novel Don Quixote not in French but in painfully accurate Spanish in the 1930s. Menard's identification with Quixote was so strong that he set himself the task of writing his favorite novel anew from his own perspective, "his aim was not to copy, but to generate the exact same text completely from scratch, as if it had never existed before"; Borges runs the two texts next to each other [which are identical], and writes, "Equally vivid is the contrast in styles. The archaic style of Menard--in the last analysis, a foreigner--suffers from a certain affectation. Not so that of his precursor, who handles easily the ordinary Spanish of his time." [!!!!!] The point here is to show "how a set of familiar words and situations, if reinterpreted in a new context, can take on utterly novel and occasionally quite bizarre auras of meaning and imagery. This theme then will become the Leitmotiv of this chapter and the following one."
281ff On rereading The Catcher in the Rye; the author reads it every decade, but this time he picked up a copy in a British bookstore and was bothered by the length of the em-dashes, certain spellings, the quote marks, etc., but it really bugged him when he saw the spellings for the word "kerb" and the use of the words "gaol" and "maths." [A type of "translation" too, and probably only strange for an American reader.] And then a pretty humorous interjection here where Hofstadter meant to quote 26 lines of The Catcher in the Rye, but J.D. Salinger didn't give him permission to use an excerpt from his book, so Hofstadter wrote 26 lines of alternative text explaining how the entire book was carefully structured so that there were page breaks in their proper place, and anything other than 26 lines would screw up the formatting of the rest of the book. So he writes precisely 26 lines to complain that Salinger wouldn't give him permission, and to explain it's so important that it was exactly 26 lines. Pretty amusing.
284 Various other existential questions here about American versus British English: does a sentence written in American English mean what it means in British English? Does a sentence written in British English mean what it means in Indian English?
286ff On the spectrum of transculturation: turning Romeo and Juliet into West side Story; transferring King Lear into a drama taking place in China with all the protagonists Chinese; or transculturation of a milder sort, like taking Melanie Mitchell's "To My Sweet" [her translation of Marot's poem] into Shakespearean England-era language, but leaves its other context undisturbed; or the various linguistic transculturations throughout Godel, Escher, Bach that happened with that work's translations into other languages. Finally on what the author calls the most local level of translation, the "atomic level"--like changing "jail" into "gaol" or "confitures" into "jam" or "gray" into "grey," etc.; the author comments on the "dubiousness" of translation at this "bottom-most rung of the ladder."
287ff The author he quotes George Steiner from his book After Babel and then makes a few interesting comments about how it's occasionally quite lyrical and makes useful observations about the subtlety of human language, but then verges into lit-crit murk that's vague, ambiguous and impossible to make sense of. [Ironically, a surprising number of writers who write about writing write like that--see for example Stanley Fish's up and down writing style in his book How to Write a Sentence; still more ironically in the next few pages even Hofstadter's style starts to become cutesy and annoying, for example using the word "biequinic" 9followed lamentably by "hee-hee" in parentheses) to refer to the French phrase deux-chevaux (two-horse, as in a two-horse carriage). It can happen to anybody I guess.]
290ff On the counterargument that "words do not have fixed imagery; context is everything" and the counter-counterargument that there is such a thing as a default image for a word or concept.
291 finally quoting George Steiner again who basically says that the Nabokov argument is "the argument from perfection": because something can't be done perfectly you shouldn't bother to do it. But yet nothing is perfect!
293ff Discussions of words that exist in some cultures but not others, or worse that exist in some era but not others: how would you translate "jazzercise" into ancient Aramaic?
295ff A couple of pages of discussion of Mandarin Chinese and its various compound word constructions; also there are a couple of interesting examples of "invisible compound words" in French given here too: like sourire [to smile] which is built from sou(s) [below] and rire [laugh] thus literally: "underlaugh."
299ff On Poul Anderson and his article "Uncleftish Beholding" which describes atomic theory but without using any Latin or Greek root words, only Anglo-Saxon root words, and thus the tone is very medieval-sounding, even though the phenomena being described is totally modern, the author uses the word "retrotranstemporation" to describe it.
302ff Finally the author writes his own transtemporation of the explanation of gravity and relative movement written by "Albert One Stone" [Albert Einstein in other words], describing gravity (or really space curvature) using "degreekified" and "dilatinified" English but also using only monosyllables [!] Note a final terrible pun here: "I have here killed two birds with One Stone."
Poems X: Struttin' my Stuff
43aff A pugilism-based translation of Marot's poem; also an interesting discussion here of unintentionally "covering your tracks": where you take a poem or a piece of writing in a certain direction, but then it takes you in another direction, and then loops you back to change what the entire work actually was, such that you jettison the one central ingredient that present in the initial genesis of that work. The author of the sky describes this as "rampant" in not just poetry translation but all sorts of creative activities. [Absolutely! Sometimes you start writing something and you realize that it has morphed into something far different from what you originally thought it would be.]
45aff "My Wee One" a translation version addressed to a very little child who's sick. And two descendent versions, "Babe o' Mine" and "Darlin' Mine" which are quite different poems with hundreds of tiny changes that causes the one poem to evolve into the next, yet the "endpoints" look radically different;
47aff A version that's sexually suggestive, and then a version version with an acrostic down the left margin. Note this comment from the author on page 48a regarding his efforts here: "As gradually I gained in skill and self-confidence, I realized that I was making new translations quite routinely and not finding the challenge nearly as hard as it had been in the beginning." So he decides to impose further constraints that would have seemed totally impossible at the outset, this is what gave rise to his acrostic version. See also a striking translation/revisioning the author does of a different French poem by Marot (this poem by Marot is called "La bonne Maguelonne" and it contains an acrostic of the Marot's full name down the left side). Hofstadter does his own version, about Jackie Kennedy Onassis, with an acrostic of his own name down the left side as well. These two works are striking.
49a Last, yet another quasi-acrostic poem here where every line ends with the letter "d." [!]
Chapter 11: Halos, Analogies, Spaces, and Blends
305ff On the extension of meanings of words via a "halo" of analogies and wider and wider usages. See the word "shadow," which has a literal meaning, but extended analogy-based meanings: a child growing up in an accomplished parent's shadow, a sick person is a shadow of their former self; thus the concept of a generalized shadow can be a source of unlimited potential analogies.
307ff On Robert L. Forward's science fiction novel Dragon's Egg, postulating a civilization on a fast spinning neutron star, analogized to Earth but on a sped up time scale. The descriptions of this civilization "work" in the story because they are analogized to Earth life. Various questions and issues discussed here with Forward's choice of words and analogies he uses. [This novel sounds like it's worth reading.]
313ff The author uses examples from the game SimTown where you have houses and goldfish in bowls and you drive in a car up and down the street, but all of these things are virtual--so do you put quote marks around these words? To what extent are they analogies? And so on.
315ff More criticisms here of Searle of using the concept of "wet" or even "alive" in a simulated/computer environment; you can't really translate this into a computer realm and consider it actual "wetness" or "aliveness" because you can't reach in and get your own hand wet, it's just a simulation; see also Stanislaw Lem and his short story Non Serviam, about life forms that have consciousness inside a computer program; John Searle also would say that because it's a simulation, there's no intentionality, or in the semantics he uses, "there's no aboutness about them." The author then asks about analogical extensions: what is the degree of similarity between the two patterns? Is the same thing happening in these two media--in the simulation and in reality? According to Hofstadter, the substrate doesn't matter: whether it's a machine, a person, a Martian's brain, whatever, what is important are "the patterns that the substrate allows to come into being."
318 Now to Robert Forward's novel, and some dialogue spoken by Cheela people on their planet (but obviously rendered in casual English): the author asks if this dialogue sounds "normal" to the reader and to what extent is it "translated" or "rendered faithfully"--because the author can't "know" the Cheelanese language because it is a fictional idea! [This problem actually makes me appreciate the science fiction work by Adrian Tchaikovsky in his Children of Time series where he renders alien dialogues in ways that seem genuinely alien. I didn't appreciate this fully at the time.] See also James Fenimore Cooper's Last of the Mohicans, where there's many pages of English text rendering conversations in Delaware, Huron, and other Indian languages; see also Pushkin in Eugene Onegin and his "translation" of Tatiana's famous letter from French into Russian, although almost certainly that was just a conceit.
320ff On other issues when you have words that get "absorbed" into other languages: see for example the word "squaw" when added to English doesn't really add anything since we already have words like woman and wife; the author says that what it adds is "flavor": this is part of the magic of the halo that can surround a word; also on other idiosyncrasies: see for example how the French translate Johann Sebastian Bach as Jean-Sébastien Bach, which makes him sound weirdly French; likewise the Poles translate Rome as "Wloch," but then why don't we translate Paris's major newspaper Le Monde as The World?
323ff Mention here of the book Mental Spaces by the French cognitive linguist Gilles Fauconnier [the author calls it "masterful"]: Fauconnier talks about how we refer to things by clearly incorrect linguistic pointers, but somehow these so-called errors actually increase the efficiency of human communication; the author then talks about related examples of what he calls "frame blends," like when someone points to their own teeth saying you have a tiny speck of tomato right "here"--clearly they're referring to the other person's tooth in that same location.
327ff Even more interesting and sometimes funny examples here: What about World War II films where all the German U-Boat officers speak English but with a German accent? [Note the even more odd and idiosyncratic example in the Tom Cruise Nazi movie Valkyrie, where most of the characters use ze German accent, but some use British accents--and of course Tom Cruise uses his typical American accent. Somehow our brains let us enjoy the movie anyway.] See other examples here in the following pages: for example when Montcalm is quoted in [the book] The Last of the Mohicans the first part of his dialogue, about one sentence, is in French, but then it switches to English with certain French words mixed in, like "mais monsieur"; also the movie The Last Emperor where the main character when he was a child spoke perfect accent-free English--which is supposed to symbolize perfect Chinese--but then when he grew up he spoke in a noticeable foreign accent in English, and yet he kept his foreign accent when speaking to his two wives, presumably in Chinese--but at the same time all of the other characters in the movie use accented English to symbolize non-English! Finally Hofstadter ties back all these examples to Marot's poem to explain why it's elegant and poetic to include French-like things in an English translation of Marot's poem: like his mother's use of "ooh la la" or other translators' use of words that evoke the same time period. All of these are examples of frame-blending.
335ff Finally returning to Robert Forward and his book Dragon's Egg again, and how both this book (which translates an alien experience into human words for human readers) is similar to translation in that there has to be frame blending consistency as well as frame blending inconsistency; it's an art, "a product of intuition and experience."
Poems XI: Hall of Mirrors
50aff A translation from one of Hofstadter's Italian friends, who made two attempts: "Mia Adorata" and "Bambina Mia"; comments here on how the meter is very flexible, the author writes "the vowels of Italian are wondrously liquid" and thus five syllable lines can be compressed into four and vice versa as necessary. [I'll also note that now that I speak some Italian I can actually read this poem, as well as compare and crosscheck the two translations that Hofstadter does--one of which is very literal and one of which takes more poetic liberties but sticks more closely to Benedetto's style and theme. But doing a crosschecking exercise like this makes me realize that when I'm reading Eugene Onegin in translation, I absolutely am not reading Eugene Onegin.] Also a discussion on his other poem "Bambina Mia" on the difference between English and Italian in the sense of their ratios of phonemes to morphemes (linguistic sounds to minimal chunks of linguistic meaning): English tends to have a little bit more "idea density" per word or per syllable than does Italian.
53aff Another translation version here in German, then a version in Russian, each with their own various idiosyncrasies, both translated into English by the author.
Chapter 12: On the Conundrums of Cascading Translation
337ff On the idea faithfulness in reproduction, including on the idea of giving readers "the illusion of actually reading foreign language poetry" even though they would be reading it in English; which means they have to overlook that the translator is adding an "extra layer of intervention."
340 The author refers back to his experience reading Falen's translation of Eugene Onegin, and thinking "Ah, Pushkin!" and noting that the translator "had managed to convince me that he didn't exist!" [Meaning the translator, because the translation was done so skillfully.]
341ff On the telephone game, and how expressions translated back and forth from one language to another take on strange characteristics; on Walter Arndt and his experiment, related in his book Pushkin Threefold, having friends translate a poem through English to French to German and then into Russian and then back to English, to see what results [a kind of "telephone by translation"]; the poems maintained their essence although they lost a lot of specifics.
347ff Here the author offers yet another analogy in geometry, where you worked on different proofs about the so-called "orthocenter" of a triangle; then more and more musings about the fundamental non-objectivity of "who gets credit" for various changes of a poem through a translation; it's sort of a tangled loop bringing up "unbelievably complex questions about what translation is."
Poems XII: Gallic Twists
55aff A translation of Marot's Old French into modern slangy, colloquial French, which the author translates back into English for us, along with a long discussion of why he decided to offer translations of Ma Mignonne into languages other than English; then the author decides to do his own modern/colloquial French version; he ends up spending an entire month on it [admittedly, there are some pretty good translingual puns throughout the poem, and various witticisms about noses, snot, blowing your nose, etc. Note these lines, late in the poem:
"Rire je n'ose" [I dare not laugh]
And he translates it into English with:
"Laugh I? Nay!"
And thus both lines end in homonyms for the other language's word for nose [nay = nez; n'ose = nose]. Incredibly impressive... if you're into this sort of thing!!
57aff On poem versions that are about the very act of translating a poem.
Chapter 13: On Shy Translators and Their Crafty, Silent Art
353ff This chapter starts out with the conversation talking about Vladimir Horowitz giving recital, one person asks what he's playing, and the other person says "I don't know, but it'll be great, Horowitz always is." This is followed by an analogical dialogue about a new translation by Gregory Rabassa of a book, where one person asks who the author is, and the other person also doesn't know, but says "It'll be great, Rabassa always is." The idea here is that you would never hear such a conversation about translating a book, but you always hear this sort of discussion about a musical performer "translating" a work of music, and yet they are more or less the same thing.
354ff Another anecdote here showing how we're often totally blind to who the translator even is: this is a reference to a spoken word version of The Iliad read by Derek Jacobi, in a way that the New York Times found very lively, apparently, but then showing a letter Joyce Carol Oates wrote to the Times about this review in which she reminded everyone that the work is in English surely, thus done by a translator [Robert Fagles in this case]. She goes on to say: "Those who imagine that foreign language works are transposed into English by a mysterious chemical process, without the effort of translators like the gifted Professor Fagles, are kin to those who imagine that film actors speak their own lines, without the benefit of screenwriters: naive." The author himself is a little ashamed, because he had sort of forgotten the same concept as well--that there was (obviously!) a translator in the middle between him and the original work, but that translator was invisible. "The only good translator is an invisible translator." in fact invisibility is the direct goal of many translators. But then the author reminds us what it would be like to go to a Horowitz performance and have a review say the performance succeeded "by disappearing from view in our listening." We would never say this about a classical music performer even though he is literally translating the works of the composer!
360ff On lip syncing: the author is mortified to learn that Deborah Kerr didn't actually sing the songs in The King and I, rather they were sung by Marni Nixon (let's hope Hofstadter never finds out about Milli Vanilli...); again, he questions "who gets credit?" Then, moving on to the idea of "explicators" versus "discoverers": see for example Martin Gardner explaining relativity by Einstein or Richard Dawkins explaining natural selection by Darwin. [It's interesting to consider this idea in light of Vox Day's concept of how "popularizers" are usually needed to explain and translate revolutionary ideas to masses.]
365 Finally the author gets to the actual point of this chapter: translators not only are totally "hidden" behind the authors, but they are "serving" those authors: they see it as an act of service to bring an author to a new audience in a different language. This despite the fact that they take tremendous liberties and their act of translation, liberties that say a performer would never take with a work of classical music for example. The author contrasts himself as a "selfish translator" who translates for the joy of creativity and the fun of taking something from one medium and transforming it with skill into another. At the same time it brings him into close contact with a work and author that he admires.
368-9 On how iterations in a poem, or a translation, have reverberations up and down the poem because you're sacrificing a rhyming word, or if affects the meter elsewhere in a line; it's like you're building a bridge, and any change you make has iterations throughout that bridge.
370 The rest of the chapter deals with how the author translated Francois Recanati's slangy "Salut, Ma Vieille" back into English, what the author calls a second-order translation. On various techniques where he flipped meanings, did reversals, etc., also discussing how sometimes you get locked into something by a choice that you've made.
377ff Interesting thoughts here on an ability to spot weak spots in a poem or in a piece of writing: the author talks about how he thinks he's written something good, but then returns to it days later and certain parts of it rub him in the wrong way. [This is definitely what happens to me with writing as well, you need to sleep on a piece of writing for at least a day, sometimes longer, and then take another look at it: the problems then become much more visible, and sometimes problems you didn't even know were there really scream out at you.] Also on the nuance of criticizing yourself versus the writing: this requires a level of detachment; also on the fact that you don't have a lot of detachment in the short run from a piece of writing, this detachment comes from time and separation.
379ff An imaginary dialogue between a computer and a human being about a translation exercise: this also doubles as sort of an extended Turing test.
Poems XIII: Pushing the Envelope
59aff Here Hofstadter asks students in one of his translation classes to take a look at some of the best or most unusual translations of Marot's "Ma Mignonne"; this stimulated them to offer some very daring (as well as some atrocious) versions of their own; also a version translated into English sonnet form in iambic pentameter, another version in a rap form.
Chapter 14: On the Untranslatable
391ff Examples here of things that are impossible to translate, like the poetry of the German poet Christian Otto Josef Wolfgang Morgenstern (1872-1914), which involves surrealism, puns and wordplay, typographic play, etc. And then on translations to English, done very well by Max Knight; Hofstadter writes "to translate something witty requires a witty translator."
394ff Quotes here on the impossibility of translating G.K. Chesterton's quote about Edward Fitzgerald's famous translation of The Rubaiyat: "It is too good a poem to be faithful to the original." A quote from Max John: "Great originals shine even through awkward translations." Also John Ciardi, an American poet and translator famous for a version of Dante's The Divine Comedy, calling translation "the art of failure," saying "What a translator tries for is no more than the best possible failure."
396ff A side by side comparison of Max Knight's Walter Arndt's translations of Morganstern, and Hofstadter tries his own attempt as well. [Another takeaway here is when you look at two or three--or in the case of Eugene Onegin, four--different translations of the work you really do get a far deeper feel for the work, you see multiple facets of it via all the translators' perspectives, and it gives you a deeply diversified view of the underlying work. Now that I think about it, when I was in undergrad I read St. Thomas More's work Utopia in one translation and hated it: it was boring, it was a grind to get through. I then read it again a few years later in a different translation and loved it. What is it about the one that made it readable and the other that made it unreadable? And which one was "faithful" to the original?If one translation is unreadable does it matter that it's more "faithful"? At the time I never thought about these things at all.]
402ff Discussion of different examples of how puns, aphorisms, slogans and even book titles can be translated: there can be a rich collection of possibilities that enable seemingly untranslatable phrases to be transferred over to another language.
404ff Examples of translating e.e. cummings and Dylan Thomas: examples the author calls "poetry at the fringes of comprehensibility" like "anyone lived in a pretty how town" or Dylan Thomas's [awful] "How Soon the Servant Son" which is pretty un-understandable in English so it would seem rather hard to translate in the first place! Even the author is willing to admit here that this brings us closer to untranslatability "as anything we have seen so far."
407ff [This chapter is losing its way, it should be cut way back]; the author retreads some of his ideas from Godel, Escher, Bach, discussing his computer program from the 1960s that produced novel-sounding English language sentences that mixed in with samples of academic gobblygook written by writers in a literary journal; the problem this book asks is how would you translate these jargon-filled pseudo intellectual sentences into, say, Chinese?
414ff On speech errors and how to translate them: the other gives an example from his book Metamagical Themas, "Rosa always date shranks" and the problems that came up with trying to represent this [unusually ungrammatical sentence] in another language. The author wanted it left done in English with an explanation, but the German and French translators made up their own examples which weren't accurate. The author sees speech errors as "pieces of data" that emerge from subconscious roots, and thus they can't really be translated, because the error will have different characteristics if copied into another language.
420ff On errors of ambiguity: see the author's hilarious [and hypnotizing] list of actual newspaper headlines from the Columbia Journalism Review:
Drunk gets nine months in violin case
Child teaching expert to speak
Milk drinkers turn to powder
Prostitutes appear to Pope
Stud tires out
Chou remains cremated
British left waffles on Falklands
Bundy beats latest date with chair
Once again there's absolutely no way to translate these into another language! [You could really only use these to teach subtle aspects of English grammar, thus they might make for a really good--albeit advanced--language class.]
426ff Pages of examples here of untranslatable things: German compound words, cultural concepts, tongue-twisters, anagrams, palindromes, etc. At the time I never thought about these things at all.
447ff Comments here on "indexical" words that depend on a speaker-dependent coordinate system or frame of reference: words like "I," "you," "here," etc., which are separate from the vast majority of words which are not seen as indexical. See for example if you did a remake of The Seven Year Itch and its self-referential joke about Marilyn Monroe as "the blonde in the kitchen" who would you use? Marilyn Monroe, or the actress's name in the remake film?
451ff Finally comments on the absolute untranslatability of this book itself on several levels.
Poems XIV: Quite Conceited
62aff Still more variations of "Ma Mignonne" but on the conceit level: a version about a cow that's quite cute and replicates the original poem's form in interesting ways; another which is a jail metaphor; thus these versions get the reader to think about different types of "sameness."
Chapter 15: On the Ununderstandable
455ff The author talks here about trip to Poland in the 1970s, where he struggled to understand what things actually cost when there were multiple exchange rates for the Polish zloty: the official one, various black market rates, etc. He uses it as an analogy for mapping one domain to another, in this case he didn't have anything fixed that he could use to map the costs or prices of things to his own framework; there was no stable exchange rate to hold onto. [Bitcoiners talk about this problem with currencies that are always diluting/debasing over time: since we measure everything by our currency, when it debases, it wrecks our ability to really "read" economic signals, it makes the world more confusing and difficult to understand.]
458 Interesting offhand comment here about how to use an expression like "Oh well, it takes all kinds" to explain someone's taste in something or to explain someone doing something inexplicable; but this is just "an easy sound bite that allows you to incorporate otherwise incomprehensible oddities into your mental framework.. .you don't really understand anything here." [I think self-sealing statements like this are a valuable clue to finding instances where we are epistemically arrogant: you want to catch these moments, you want to unearth them, so they don't trip you up somehow.]
459ff Carious examples of taste in music and climbing into the mind of someone who has much different tastes than you do. [This chapter also quickly loses its way, it has very little to do so far with the book; note for examle on page 463 where the author, strangely, likens "dislike for jazz" to have roots in Nazism for no apparent reason at all. See also the even stranger footnote for page 463 (which is located on page 593) where the author talks about a book that he thought was going to criticize rock and roll, but it wasn't the book he thought it would be, but yet made him think about a book he would like to see written one day. This is one of the strangest, least scrutable footnotes I've ever read.]
465ff Page after page of analogies here that the author uses to attempt to understand why someone who was a member of a fanatic sect would kill a priest.
469ff Multiple long quotes from Ean S. Connell's book Diary of a Rapist, on the idea of trying to understand the mentality of the book's narrator.
474ff Very long multi-page discussion about who one identifies with in the movie The Seven Year Itch: if you're a man you'd identify with the Richard Sherman character, but what about if you're a woman? Do you identify with the wife who's up in Maine? Or the other woman?
481ff On the "I": what is it when a person dies, what survives; what about the "you" who you were 10 years ago, where is that? The author argues "that person is just a set of memories inside your brain (and those of other people.)" References here to his other book The Mind's I: Fantasies and Reflections on Self and Soul, coedited with Daniel Dennett; on the survival of a soul in other brains via memories of that person. Also citing the book Reasons and Persons by Derek Parfit, who writes on these subjects.
485ff On relating to one's former self or former selves: something the author considers as complex and confusing as relating to other people. [Unless you're a narcissist, then other selves don't even exist for you in a meaningful way, except for narcissistic supply...]
486ff This chapter is really wandering. The author considers the nature of empathy, considers the death of a language and what that would mean, then starts dwelling on artificial intelligence, and then begins talking about the chess match between Gary Kasparov and Deep Blue. Also a discussion on if true human-thinking robot computers ever roamed the Earth would they be benevolent or friendly, this is again way off topic. This chapter and the last could have been cut or compressed down considerably.
Poems XIV: That Ol' Sino Room
65aff The author is asked if he ever ran "Ma Mignonne" through a machine translation program? His immediate response is "What for?" as he presumed the translation would be atrocious. But he tries it anyway, and finds that the translation is at the level of total literality: ignoring form, line breaks, rhymes, syllable counts; the output read like it was a washing machine repair manual. The point here is the machine translating programs don't "understand" the material in the least. Even the very advanced machine translation done in 1992 by IBM's Candide program still sucks, even though it has access to tremendously gigantic databases of translation contexts. [Note that Google translate today is far better than any of these programs, although it's still fails miserably with poetry and metaphorical speech.] The Candide program translated "Ma Mignonne" as "My Flapper," based on language context about the 20th century; the bulk of the database references for the word "mignonne" were used to describe a 1920's-era flapper girl; mignonne was also a word used to mean flapper. Obviously in the context of the poem this is a catastrophic mistake.
68aff A version here by a Chinese person who knew very little English and no French, but who was armed with a bunch of dictionaries: she attempts a translation, and it comes out at about the same quality--and with the same types of problems--as the machine translators. In other words what was missing, catastrophically, was context.
Chapter 16: AI Aims, MT Claims, Sino-room Flames
495 Different examples here of using words, terms and expressions with varying levels of "knowing": a baby asking for a "pacifier" but not knowing it is to be "pacified"; using technical terms that you've just learned as if you're very familiar with them; also or an example of a French-speaking person using baseball idioms in English perfectly, etc.; the point here is that we can just parrot words, and while it sounds like we "know" the words we're using, it doesn't necessarily mean that we do. [I'm reminded of terms like "QE" or "blockchain technology" other trendy terms that show up in the media-propaganda machine that people will parrot in normal parlance as if they really know what the terms mean.] Basically the idea here is to understand the difference between parroting a term and actual semantic mastery of that term.
498ff On "aboutness": when computers "use" words, those words are "about" nothing, but humans can incorporate context, meaning and understanding into the words they use; in contrast, symbols in a computer have an "aboutlessness" about them.
500ff Various problems of context loss and contextlessness of terms with machine translation; on controlling for it by limiting the domain of discourse, see for example an extremely limited topic like a copy machine repair manual, or writing in a very limited style; see Xerox's "Multinational Customized English."
504ff On the difference between semantics and syntax; complicated by the computer world's overbroad use of syntax as programming/computing terms.
508ff On SHRDLU, a program we met in GEB, that has some ability to handle semantics. Hofstadter claims that there is "aboutness" in this program. Also on the Proteus program that appears to have aboutness in the microdomain of tic-tac-toe as it both plays and runs commentary and analysis.
515ff On driving as another aspect of AI [note that this book was written in the 1990s so the author could have no idea about Waymo]; for now he considers tasks like driving to be way beyond anything AI can do. [This is surprising, given that Hofstadter thinks AI has enormous potential--even the potential for self-awareness.] Also comments here on what "driving" means: one [very limited] meaning is to keep your car on the road, but also it means many, many things beyond just that task.
520ff On smoke screens that are thrown up on both sides of the AI debate: John Searle's argument that there's no "aboutness"; that these are like can openers; or the smoke screen of all the unjustified hype about AI that produces an inevitable backlash that discredits the domain as well.
Poems XVI: My Sweet Ones
70aff These are very personal and personalized translations of Marot's poem: one version that the author wrote for his mother after she had cardiac surgery; another other for his wife when she had surgery, and then finally one that his wife wrote for him. These are all quite beautiful and touching variations.
Chapter 17: In Praise of the Music of Language
523ff Here the author attempts to define poetry, as well as make an argument that poetry needs to have a pattern--or have some sort of aesthetically restricted medium--in order to be poetry. He doesn't quite it this way, but he talks about a sort of "inversion" of poetry in the modern and postmodern era, where poetry became too sophisticated for its own good: where structures like rhyme and rhythmicality became denigrated as too "easy" or too "populist" and thus obscurantist and esoteric freeform verse became all the rage, the author considers poetry to have become elitist and pompous. [He's right!]
528ff On looking at Dante's Divine Comedy and translations by several different English-speaking poets; on the structure of the Divine Ccomedy in Italian: the 11 syllable lines, the triplet rhyming and the triple rhyme/terza rima which puts somewhat frightening constraints on the translator; it's kind of interesting to hear his take on the work itself, since he's not particularly Christian and didn't live in Northern Italy during the 13th and 14th centuries, he makes a remark about how hell is mostly populated by Northern Italians so it seems like a small town [heh]; he likens the inter-city rivalries of the region to the rivalry between Stanford and Berkeley.
532ff On Robert Pinsky's 1994 translation of Dante's Inferno, and on Pinsky's choice for a looser definition of rhyme in English, using a "Yeatsian rhyme" where words sound similar but don't rhyme exactly. The author's mind is "boggled" by Pinsky's decision to "tamper with the very core of the structure"; he's offended that he would mess so radically with Dante's rhyme structure.
535ff Hofstadter goes on to criticize Seamus Heaney and his version of the first four tersets of The Inferno, finding it "extremely depressing."
537ff And then he criticizes the translator Mark Musa for "poverty of imagination," as Musa claims that doing both rhyme and meter makes it impossible, therefore it can't be done.
538 Note the story here about Willis Barnstone, who spent time with Jose Luis Borges in the 1970s, working on translating his sonnets into English; Borges read one of his works and told him to "try a little harder" with some of the rhyme choices.
541 Onto more direct translations of The Divine Comedy by Dorothy Sayers and Richard Wilbur, both of which are more in the direct style of Dante himself, as opposed to the highly tortured syntax of translations by Melville Best Anderson and Laurence Binyon. But he gives credit to all four for maintaining the terza rima structure.
545 More versions of Dante, including translations by Daniel Halpern, Cynthia Macdonald and others, all of which either punted the rhyme structure totally or were "wishy-washy" about it, followed by a genuinely funny comment here where Hofstadter describes himself as "the Vladimir Nabokov of The Inferno" as he rants about how these translations disappoint him.
548 More here on the author being the "Vladimir Nabokov of Dante" as he comments that while Nabokov believed that with Pushkin you could not translate poetry as poetry (and that was Nabokov's "religion" regarding Eugene Onegin), Hofstadter just as strongly feels that the translator has to have "a reverence for pattern" and that is his religion.
550ff On having a good sense of humor: being able to know other people's expectations, and thus know how to violate those expectations; how to play with language, how to hear strange connections and phonetic resonances, how these are traits of both intellectual risk-taking, as well as in poetry and poetry translation.
553ff A section here on translating Basho's famous haiku:
Furike ya
kawazu tobikomu
mizu no oto.
Translated by Hiroaki Sato into a single line, with a single thought, with absolutely no excess verbiage:
An old pond: a frog jumps in -- the sound of water.
Followed by many other versions, some of which maintain the syllabic and line structure of a traditional haiku to various degrees.
557 "Art must be rendered as art, otherwise it is no longer art." This is a one sentence explanation of the author's philosophy of translation: you can't just take poetry and translate it as unrhymed non-metric prose. "Content and form must both be preserved."
Conclusion: Le Tombeau de ma rose
560ff The author talks about some of the translations he tried to do of the songs performed at his wedding, and then some attempts he made later to retranslate them, then a discussion of his wife experiencing severe headaches after they moved to Italy; on her rapid decline as a tumor was discovered in her brain. The book ends in a very sad, bittersweet way, it's beautiful.
Vocab:
Convolving: to roll or twist together; can refer to a physical action, or, more commonly, a mathematical operation that combines two functions
Eleemosynary: related to or supported by charity
Syzygy: the alignment of three celestial bodies in a straight line
Oligoglot: someone fluent in a few languages (Oligo = few, as in oligarchy, glot = from the Greek glotta, tongue, having/speaking a language)
Metonymy: the substitution of the name of an attribute or adjunct for that of the thing meant; "suit" for business executive
Morpheme: a meaningful (morphological) unit of a language that cannot be further divided
Echt: authentic and typical
Opuscule: book; a small or minor literary or musical work
Osculate: to kiss; [math]: on the touch of one curve or surface with another so as to have a common tangent at the point of contact
Poietautonym: a poet's self-reference [I think this is a Hofstadter neologisms]
Lipogram: text written under the constraint that one or more letters are not allowed to be used
Yclept: by the name of ["Peking University is oddly retro-yclept"]
Termagant: a harsh-tempered or overbearing woman; harpy
Antiphrasis: a figure of speech where a word is used in a sense opposite to its usual meaning, often for ironic effect; calling a very large person "Tiny"
Enjambment: the continuation of a sentence or an idea from one line or couplet of a poem to the next
Amphisbaenic: (Greek, lit: "go both ways") like a mythical serpent with a head at each end
Defowlerize: a neologism coined by Vladimir Nabokov (which is a dig at Henri Watson Fowler’s style guide, Modern English Usage) meaning to use a non-normative English style [This is a great word! You can also see the word "deflower" in here as well, "deflowered English" is a pretty powerful way to think about it]
-Ast/-astry: [suffix, latinate] inferior, not genuine; something that imperfectly resembles or mimics the true thing [e.g.: paraphrast, poetastry]
Ictus: [Prosody] a rhythmical or metrical stress; [Medicine] a stroke or seizure; a fit
Media:
Music of Alberto Rabagliati, Guiseppe Lugo, Lina Termini, Luciana Dolliver, Ernesto Bonino
Music of Gabriel Fauré ("Aurore," "Notre Amour")
Louis Armstrong: "St. James Infirmary"
Silent Movie [Mel Brooks film]
John McCormack performing "I'll Walk Beside You"
To Read [comments in quotes are from the author]:
***Willis Barnstone: The Poetics of Translation ["witty"]
***Vikram Seth: The Golden Gate
Christian Morgenstern: Galgenlieder (trans. Max Knight)
***Omar Khayyam: The Rubayat (trans. Edward FitzGerald)
Ronald Storrs: Ad Pyrrham
Julian Offroy de la Mettrie: L'homme Machine
***Pamela McCorduck: Machines Who Think
Charles Babbage: Passages from the Life of a Philosopher
***George Steiner: After Babel
Warren Weaver: Alice in Many Tongues
Mikhail Bongard: Pattern Recognition
***Joseph Weizenbaum: Computer Power and Human Reason
***Charles Kay Ogden and I.A. Richards: The Meaning of Meaning
Edgar Allan Poe: "The Philosophy of Composition" [essay]
***Luo Guanzhong: Romance of the Three Kingdoms [unabridged, four volumes; trans. Moss Roberts]
George Steiner: Fields of Force: Fischer and Spassky at Reykjavik
***Raymond Queneau: Exercises In Style (trans. Barbara Wright)
Alexander Pushkin: Eugene Onegin (trans. Charles Johnston)
Vladimir Nabokov: Strong Opinions
***Walter Arndt: Pushkin Threefold
***Robert L. Forward: Dragon's Egg
***Gilles Fauconnier: Mental Spaces ["masterful"]
Andrew Hodges: Alan Turing: The Enigma ["definitive"]
***Jean Aitchison: Words in the Mind ["one of my favorite books on how the mind works"]
Derek Parfait: Reasons and Persons ["marvelous though difficult"]
Eva Hoffman: Lost in Translation ["poetic and eloquent"]
Hiroaki Sato and Burton Watson: From the Country of Eight Islands [translations of 1200 years of Japanese poetry]